Skip to main content

Full text of "Journal of the Society of Motion Picture and Television Engineers"

See other formats

From the collection of the 

x. z m * 


1 v Jjibrary 

t p 

San Francisco, California 

Journal of the Society of 

Motion Picture and Television Engineers 


Report of the President EARL I. SPONABLE 3 

Perception of Television Random Noise PIERRE MERTZ 8 

An Improved Photomultiplier Tube Color Densitometer 


Color Measurement of Motion Picture Screen Illumination 


Cinecolor Three-Color Process ALAN M. GUNDELFINGER 74 

New Projection Lamp and Carbon-Feed Mechanism. . J. K. ELDERKIN 87 

Industrial Sapphire in Motion Picture Equipment 


Report of SMPE Standards Committee FRANK E. CARLSON 102 

New American Standards 106 

Z22.61-1949, American Standard Sound Focusing Test Film for 35- 
millimeter Motion Picture Sound Reproducers (Service Type) . . . 107 
Z22.68-1949, American Standard Buzz-Track Test Film for 35-milli- 
meter Motion Picture Sound Reproducers 108 

New Officers of the Society 109 

Engineering Committees Ill 

1950 Nominations 113 

Society Awards for 1950 113 

Society Announcements 115 


Acoustic Measurements, by Leo L. Beranek. Reviewed by Harvey Fletcher . 117 

Painting with Light, by John Alton. Reviewed by John W. Boyle 118 

Feininger on Photography, by Andreas Feininger 

Reviewed by Lloyd E. Varden 118 

The Complete Projectionist, by R. Howard Cricks 

Reviewed by Merle Chamberlm 119 

Current Literature 120 

New Products 121 

Meetings of Other Societies 123 


Chairman Editor Chairman 

Board of Editors Papers Committee 

Subscription to nonmembers, $12.50 per annum; to members, $6.25 per annum, included in 
their annual membership dues; single copies, $1.50. Order from the Society's General Office. 
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and 
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture Engineers, 
Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General and Editorial Office, 
342 Madison Ave., New York 17, N. Y. Entered as second-class matter January 15, 1930, 
at the Post Office at Easton, Pa., under the Act of March 3, 1879. 

Copyright, 1950, by the Society of Motion Picture Engineers, Inc. Permission to republish 
material from the JOURNAL must be obtained in writing from the General Office of the Society. 
Copyright under International Copyright Convention and Pan-American Convention. The 
Society is not responsible for statements of authors or contributors. 

Society of 

Motion Picture and Television Engineers 

342 MADISON AVENUE, NEW YORK 17, N. Y., TEL. Mu 2-2185 

Earl I. Sponable 
460 W. 54 St. 
New York, 19, N.Y. 

Peter Mole 

941 N. Sycamore Ave. 
Hollywood 38, Calif. 



Loren L. Ryder 
5451 Marathon St. 
Hollywood 38, Calif. 

Clyde R. Keith 
120 Broadway 
New York 5, N.Y. 

Robert M. Corbin 
343 State St. 
Rochester 4, N.Y. 

William C. Kunzmann 
Box 6087 
Cleveland 1, Ohio 

Fred T. Bowditch 
Box 6087 
Cleveland 1, Ohio 


Frank E. Cahill, Jr. 
321 W. 44 St. 
New York 18, N.Y. 



Ralph B. Austrian 

25 W. 54 St. 

New York 19, N.Y. 


Herbert Barnett 
Manville Lane 
Pleasantville, N.Y. 

Kenneth F, Morgan 
6601 Romaine St. 
Los Angeles 38, Calif. 

Norwood L. Simmons 
6706 Santa Monica Blvd. 
Hollywood 38, Calif. 


George W. Colburn 
164 N. Wacker Dr. 
Chicago 6, 111. 

Charles R. Daily 
113 N. Laurel Ave. 
Los Angeles 36. Calif. 

John P. Livadary 
4034 Cromwell Ave. 
Los Angeles, Calif. 

Edward Schmidt 
304 E. 44 St. 
New York 17, N.Y. 


Lorin D. Grignon 
1427 Warnall Ave. 
Los Angeles 24, Calif. 

Paul J. Larsen 
508 S. Tulane St. 
Albuquerque, N.M. 

William H. Rivers 
342 Madison Ave. 
New York 17, N.Y. 

Edward S. Seeley 
161 Sixth Ave. 
New York 13, N.Y. 

R. T. Van Niman 
4501 Washington St. 
Chicago 24, 111. 

Report of the President 



IT is CUSTOMARY at this time for the President of the Society to sub- 
mit his annual report, and to bring to your attention such matters 
as seem to him worthy of your consideration. First, I am glad to be 
able to report to you that the year 1949 has been filled with healthy 
activity on the part of the Society and has resulted hi an even greater 
service to the motion picture industry than at any time in the past. 

The total membership at this time is over 3000: a real credit to 
the Membership Committee which has brought in 280 new members 
since April 1 of this year. This number of new members has more 
than made up for an unusually high loss of previous members, due 
probably to the aftermath of the war and to changing business con- 
ditions. Among our membership are representatives from 48 foreign 
countries, including Canada and Mexico, and from each of our 48 

I am proud of the work done by our 38 standing committees in which 
471 members are giving their time to help improve the industry 
through standardization, and in other important ways. Parentheti- 
cally, at this point I would like to say that we would welcome the 
assistance of any members who are not now serving on committees, 
and who would like to do so. I suggest they get in touch with the 
chairman of the committee in which they are interested. 

The Officers and Governors have been diligent in their jobs and 
most patient and helpful to me at Board Meetings. I want par- 
ticularly to thank those men who are now finishing their terms of 
office and to urge them not to diminish their activity in Society 

The general office is now well established at 342 Madison Avenue 
in New York City. The space, while not elaborate, has been adequate 
for our present needs. By the way of review, we have an efficient, 
small staff headed by Boyce Nemec, our Executive Secretary. The 
work of the engineering committees is handled by William H. Deacy, 
Jr., Staff Engineer; and Sigmund Muskat is the Office Manager. 

PRESENTED: October 10, 1949, at the SMPE Convention in Hollywood. 




Improvement has been made in various office activities : The account- 
ing system has been modernized; the committee work is being handled 
rapidly and efficiently; membership applications are being processed 
promptly; Journal publication has been speeded up; and our press 
relations are much improved. 

Besides its own committee work, the Society now contributes to 
the support of, and has representation in, related organizations in- 
cluding the American Standards Association, the Inter-Society Color 













Council, the United States National Committee of the International 
Commission on Illumination, the American Documentation Institute, 
and the National Fire Protection Association. 

Our three local Sections have been unusually active during the 
current year. The chairmen and managers have arranged for papers 
that have commanded increased attendance. One highlight of the 
year was the jioint meeting of the New York and Chicago Sections 
through the use of inter-city television, dealing with "A Study of 
Television Lighting." The combined attendance at this one meeting 
exceeded 1000 and resulted in important publicity and improved 
public relations for the Society. 


The 65th Semiannual Convention, held this past spring in New 
York City, was an outstanding success. Its theme of television at- 
tracted an all-time record registration of 715. This current Conven- 
tion, the Society's 66th, also has a program of great interest that has 
been made possible by most diligent work on the part of the Program 
and Local Arrangements Committees. I am sure that this Hollywood 
Convention will be one that we all shall remember as an enlightening 

Printing costs have gone up along with everything else. This has 
led to a study of the format of the JOURNAL, and designing it to use 
the available space more efficiently. Careful planning is in progress 
in co-operation with the Society's printer to effect a transition to a 
somewhat new dress wherein we may use for the most part a two- 
column page and achieve better readability and more text material 
per page, at practically no increase in cost. 

During the year the Society published a book on theater engineering 
entitled The Motion Picture Theater, with which I believe most of you 
are familiar. In spite of a carefully worked out plan for sales we 
have not received the number of orders we anticipated. I believe, 
however, that an important reference book such as this will be in 
demand for some time to come. Our other less ambitious publications 
in the form of monographs entitled "Films in Television," " Theater 
Television," "High-Speed Photography," and "Color Symposium" 
were well received, have paid their costs, and have helped to gain 
the Society recognition in their respective fields. 

The Society has continued its service of making and supplying 
test films. New films this year include a 16-mm sound service test 
film, and a television visual test film. Both of these films are very 
much needed in the industry. They will be described and samples 
will be shown at this meeting. 

A number of new standards have been approved during the current 
year, and have been published in the JOURNAL. Also, the Board of 
Governors has authorized the publication from time to time of special 
reports to be known as "SMPE Recommendations." While these 
Recommendations have not reached the stage of standards, they will 
have been approved by engineering committees and the Engineering 
Vice-President and should be a useful guide to equipment manufac- 
turers. They will be printed to fit the standards cover, but colored 
paper will be used to distinguish them from adopted standards. 


The financial position of the Society, while on a sound basis at 
present, requires careful watching. Our expenditures and receipts 
run about $125,000 per year. The year 1948 ended with a deficit 
of $8,724. This was partly due to non-repetitive expenses such as 
furniture for new offices. This year the indications are that we will 
nearly break even. Our net quick assets are over $90,000. Largely 
through the efforts of Don Hyndman we have materially increased 
our income from sustaining members. In 1945, income from this 
source was $8,087, was brought up to $20,250 in 1946, and this year 
it will total over $24,000. If we are to continue to expand our service 
to the industry it is obvious that our income must keep in step. 
Every possible source of revenue will be studied carefully and every 
effort made to keep our operating costs as low as is consistent with our 
program of service. 

Never in the history of this Society has it had the standing, or 
commanded the respect of the leaders in the motion picture industry, 
that it does today. This is no mere accident, but rather is the cumu- 
lative result of teamwork among all its members. The pioneer work in 
theater television, largely due to the efforts of Paul Larsen and a few 
others, is beginning to be recognized. The recent Society answer 
to the request of the Federal Communications Commission for advice 
regarding theater television (which will be reported on later at this 
meeting by Don Hyndman) is an example of what I mean. Like- 
wise, the Theatre Owners of America and the Motion Picture Asso- 
ciation have sought our advice in this same matter. We must carry 
on and justify the confidence that has been placed with us to do the 
proper engineering job and to give technical guidance, not only to 
the motion picture industry as it is now constituted, but also to the 
new, closely allied art of television. 

This brings me to our plans for the future. Our Past President, 
Loren Ryder, has emphasized over and over that the scope of the 
Society activities includes all phases of pictorial rendition of action. 
With this I am in hearty accord. We are concerned with television 
whether we all like it or not. While television did not develop within 
the motion picture industry and while credit for its conception and 
growth belongs to the electronic and radio engineers, there is, never- 
theless, a very large area of common interest in the two fields. The 
vast accumulation of knowledge of production problems, lighting, 
photography, sound recording, film handling, and projection tech- 
nique are all parts of this common area; and this accumulated knowl- 


edge of the membership of this Society represents such values to the 
growing television art that the televiiion engineer will find himself 
able to acquire this information in only one of two ways: either by 
arduous, costly, personal experience, or alternatively, by becoming a 
member of this Society. We have, therefore, much incentive to 
offer to the television engineer to join with us; and on our side, there 
is much to be gained by this union, both from the point of view of 
society economics and from that of service to the industry. It is 
for these basic reasons that your Board, after due committee consid- 
eration, decided to recommend to the membership that the name 
of the Society be changed to "Society of Motion Picture and Tele- 
vision Engineers," and that the founders and developers of this new 
allied art be actively encouraged to take part with us in developing a 
larger and more effective service. It is my sincere personal belief 
that such a change will profit the Society and the industry, and I 
hope that with your enthusiastic support of the enlarged program 
which I have just outlined, time will prove the wisdom of this course. 

Perception of 

Television Random Noise 



Summary The perception of random noise in television has been clari- 
fied by studying its analogy to graininess in photography. In a television 
image the individual random noise grains are assumed analogous to photo- 
graphic grains. Effective random noise power is obtained by cumulating 
and weighting actual noise powers over the video frequencies with a weight- 
ing function diminishing from unity toward increasing frequencies. These 
check reasonably well with preliminary experiments. The paper includes 
an analysis of the effect of changing the tone rendering and contrast of the 
television image. 

THE ACCUMULATION of data for guidance on tolerances to be placed 
on random noise in circuits used for television transmission has 
developed a fairly large number of parameters that cannot be neg- 
lected if the interpretation of these data is to be useful. 1 Among 
these parameters are some which concern the phenomena involved in 
the perception of the noise by the viewer. There is presented here- 
with a discussion of some of these phenomena. 

The treatment covers several of the parameters, but it cannot pre- 
sume to solve completely the problem of tolerances. It constitutes 
merely a first order attack on the major quantities involved. 

The discussion is divided into three parts: (1) factors involving 
the granular appearance of the random noise; (2) factors involving 
the perception of adjacent differences in luminance; and (3) factors 
involving the translation of signal voltages into image luminances. 


Certain phenomena in the perception of random noise have been 
clarified by studying its analogy to the effect of graininess in a photo- 
graphic image. 

Long study of photographic graininess 2 indicates that the percep- 
tion of graininess involves two parameters of the emulsion : 

1. The extent of the density variations in the emulsion caused by 
the grains; 

2. The average size of the individual grains. 

PRESENTED: October 14, 1949, at the SMPE Convention in Hollywood. 



a = a constant 

A = a constant, in one case being the constant area of a microdensitometer 


b = a constant, in one case being a characterization of the "key" of a pic- 


B = luminance, average picture luminance over an area 

BA = adaptation luminance (millilamberts) 

BB = luminance of test field (millilamberts) 

Bs = maximum luminance in surround field (millilamberts) 

B(6> <P) luminance of surround field over elementary solid angle rfw 

c = a constant 

D = ratio of actual viewing distance to a picture, to four times picture 


/ = frequency 

/, = effective frequency 

f = upper cutoff frequency of low pass filter 

g = quantity evaluating appearance of graininess 

= proportionality constant 

K a constant, in one case a constant speed of scanning beam 

n = a number, in one case used as exponent, in another, an average number 

of photographic grains per unit area 

N = number of television scanning lines in picture height 

p, q = constants 

r = response of eye to granular luminance deviations 

s = index of graininess in photographic emulsion 

S = sensation (evaluated in number of perceptible steps) 

T = time taken by scanning or reproducing beam in sweeping across as- 

sumed sampling area when Z> = 1 

u = number of television scanning lines in sampling area height 

v = average number of noise spots per segment of scanning line across 

sampling area 
= signal voltage 
= mean square of luminance deviations (or "power") of random "noise" 

W e = effective square of luminance deviations (or "power") of random 

Wi = "power" at threshold for "flat noise" 

W z = "power" at threshold for "up-tilted noise" 

W(f) = mean square of luminance deviations (or "power") of random "noise" 
in unit frequency band at frequency f 

8B = rms of luminance departures from average, in a portion of the picture 

A& = increment in luminance, measured in effective photographic density 

AB = increment in luminance, measured in millilamberts 

Ay = increment in signal, measured in decibels 

A V = increment in signal, measured in volts 

<t> = angle measured about line of sight to a test field 

8 = angle in radians between line of sight to an elementary spot d and 

Hne of sight to a test field 

<r\ = rms departure from average in microdensitometer density measurement 

(idealized aperture) 

<TJ = rms departure from average in microdensitometer density measure- 

ment (actual aperture) 

<fa = elementary solid angle in steradians 

The graininess of a photographic emulsion is measured by exploring 
an arbitrary path over a region of it having constant average density 
over the gross parts of the region. This is carried out with a micro- 
densitometer having a sampling aperture which is small but of suffi- 

10 PIERRE MERTZ January 

cient area to include a number of grain clumps simultaneously. From 
the microdensitometer record there is determined the root mean 
square deviation of the density about the mean. 

A highly schematic illustration of this process is shown in Fig. 1. 
The sampling aperture is shown in one position along the path. It is 
shown square rather than with the usual round form to simplify some 
later discussion. 

The microdensitometer readings will vary according to the size of 
the aperture used. In Fig. 2 an illustrative trace of the density is 




Fig. 1. Scheme of microdensitometer sampling. 

cr, (R M s) 





Fig. 2. Example of density readings. 

compared with an idealized measurement made with an infinitesimal 
aperture. The ratio of the rms departures (<TI and <r 2 ) in density is ap- 
proximately equal to the square root of the number of grains included 
in the aperture, thus 


where n = average number of grains per unit area; 

A* = area of effective aperture, measured on emulsion. 

Hence in correlating measurements made with different apertures 
on the same emulsion, a*constant quantity is the product of the rms 
density measurement by the square root of the aperture area. This is 
indicated as equal to the idealized *rms density measurement divided 
by the square root of the number of grains per unit area. 

s = <n/\/n = <7 2 vGl (2) 


This quantity is here called s, and used as an index of the graininess 
of the emulsion. 

When the emulsion is viewed by the eye, a less definite but similar 
sampling aperture comes into play. This is caused by the limited 
resolving power of the eye, instead of the microdensitometer. The 
appearance of graininess, or evaluation by the eye of the quantity <r 2 , 
is then equal to the quantity s divided by the square root of the samp- 
ling area. This new quantity is called g. 

g = *2 = s/A (3) 

The sampling area subtends a constant solid angle at the observer's 
eye as the viewing distance is changed. Thus at the greater viewing 
distances the area A on the emulsion increases, and the appearance of 
graininess is reduced. 

In a television image the individual random noise grains are as- 
sumed as analogous to the photographic grains. The sampling area 
again subtends a constant solid angle at the observer's eye. The 
number of noise grains in this area is proportional to the product of the 
number of scanning lines in the area by the 
average number of grains in the portion of a 
scanning line included in the area. This last 
number can be computed if desired from a 
formula published by S. 0. Rice 3 for the average 
number of zeros per second in random noise. ( _ } u L|NES 

It is found, however, that in television there 
is correlation between noise grains along a 
scanning line, according to the particular noise 
power distribution in the frequency spectrum. 

This is a fact which does not figure in the Fi s- 3. Television 

... sampling area. 

photographic analogy. In consequence the 

analogous appearance of graininess for the television must be com- 
puted in a slightly different manner. 

A sampling area is illustrated in Fig. 3. A distribution of lumi- 
nances along one scanning line is illustrated at the top. For small de- 
partures the deviation in luminance is nearly proportional to the nega- 
tive of the deviation in density. The average response of a photocell 
over such a sampling interval of scanning line is known from scanning 
theory, 4 and is taken as analogous for the eye. Each Fourier compo- 
nent in the trace is attenuated by a weighting function characteristic 
of the sampling aperture or interval. In this case the sampling inter- 

12 PIERRE MERTZ January 

val will be assumed rectangular, of duration DT. D is a factor weight- 
ing viewing distance, which at four times picture height is taken equal 
to one, i.e., 

n - viewing distance 
" 4 X picture height ' 

T represents the tune taken by the reproducing beam in sweeping 
across the sampling area when D = 1. The width of the sampling 
area on the picture screen is KDT, where K is the speed of the beam. 

The "power" response, using this in the sense merely of the output 
of the photocell analogous to the eye, of a single scanning line trace, is 
given by 

ri = / W(f) [sin irfDT/^fDT)]* df. (4) 


The correlation which exists between noise grains along the scan- 
ning line in the sampling area drops to zero between scanning lines. 
Thus the square of the rms luminance deviation (or the "response 
power") averaged over the number u of scanning lines in the sampling 
area height is, as in the photographic case, the square of the rms de- 
viation (or "response power") over one line divided by u or kDN. 

Here N = number of scanning lines in picture height; 
k = proportionality constant. 

With all the constants adjusted to give 1 at D = 1, the effective noise 
power response is given by . 

W e = rVD == (l/D)J W(f) [sin TT fDT '/ '( V /DT)] 2 df. (5) 

Thus the effective random noise power is obtained by weighting 
and cumulating actual noise powers at the various video frequencies 
with a weighting function. This function diminishes from unity to- 
ward increasing frequencies approximately like the weighting func- 
tion of a scanning aperture. Following this theory, then, one would 
expect threshold of perception to be obtained at a fixed effective ran- 
dom noise power for all distributions of the random noise. 

The theory can be checked with some preliminary unpublished ex- 
perimental data taken by M. W. Baldwin on the near threshold values 
of various distributions of television random noise, viewed at several 
distances. The twelve distributions experimented with are illustrated 




in Fig. 4. "Flat" means a distribution which is substantially flat up 
to cutoff. "Up-tilted" means one in which rms amplitude in a narrow 
frequency band is proportional to the center frequency of that band 
up to the region of cutoff. "Coaxial" is a distribution which is ex- 
pected in some hypothetical coaxial system designs. These noise dis- 
tributions were viewed at distances D = 0.625, 1.0, and 2.0, respec- 
tively. The near threshold values of total measured random noise in 
each case as a function of upper cutoff frequency are indicated by the 
connected points in Fig. 5. The noise is measured in terms of the 
ratio of the rms of luminance departures 5B, to the average luminance 














I 23456 I 23456 I 23456 


Fig. 4. Experimental random noise distributions. 

B. The logarithm of this is taken, as if db were used to represent an 
amplitude ratio. The quantity has sometimes been called "decil- 

In fitting equation (5) to these data it is necessary to choose a value 
for the sampling interval T. As will be indicated further below, this 
has been selected for the best fit at T = 0.22 microsecond. With this 
value the weighting functions for the three distances are shown by the 
curved lines in Fig. 6. The fit of the computation of equation (5) with 
experimental data under these conditions is shown by the curved 
lines in Fig. 5. 

The fit is not perfect, but is considered reasonably good. The 
greatest discrepancies in trend seem to come at the shortest viewing 
distance. This suggests that the noise distributions which have been 


















to u> r 
o o t 














* ' 



K^ 5 



~X^2 3 4 5 ( 

/ 2 3 4 5 








/ i 




i / 













/ 2 3456 

Fig. 5. Noise perception with distributions of Fig. 4; comparison with theory. 


Fig. 6. Hypothetical visual weighting of noise. 

shown in Fig. 4 are modified by the filtering effect of the picture tube 
electron beam spot before being viewed by the observer. The exact 
distribution of luminance in the spot is not known, but a computation 
of the filtering effect is shown in Fig. 7 on the assumption that the dis- 
tribution follows a cosine squared law of such width as to come to zero 
weighting at 10 Me. 





o _. 

I 2 


5 6 

Fig. 7. Hypothetical filter effect of receiving cathode ray spot. 











3 ^ 





D = 


p ,^X' 








' X 















- -* 














Fig. 8. Noise perception with distributions of Fig. 4; comparison 
with theory corrected by Fig. 7. 

With this correction the experimental data are fitted as shown in 
Fig. 8. The fit is somewhat better than in Fig. 5, and the worst resid- 
ual discrepancy is reduced to 3 db. 

It would be expected that the law derived from analogy with photo- 
graphic graininess should break down at very low frequencies because 

16 PIERRE MERTZ January 

the granularity in the horizontal direction then extends well beyond 
any reasonable sampling area. Such a breakdown has been found to 
exist in the case of narrow band distributions of noise well under 1 Me. 
The low-frequency region and distributions under which the law ap- 
pears invalid seem narrow enough not to be of too much consequence 
in its general use. 

There is some interest in returning to the evaluation of the response 
from the sampling area of Fig. 3 and disregarding the correlation be- 
tween noise grains along a scanning line. As noted above this can be 
done from Rice's treatment. 3 It can be assumed for the present pur- 
pose that the noise has an "effective frequency" f e , which is defined as 
half the number of zeros per second. 




W(f) df 

Thus the effective frequency is the radius of gyration of the power 
distribution figure about the y axis. The average number of spots of 
noise v per scanning line in the sampling area is equal to : 

v = DTf. . (7) 

Thus the total average number of noise grains in the sampling area 
is equal to uv. Hence following equation (1), but writing powers in- 
stead of density departures, and placing na = uv, 

W, = W/(uv) 

= W/(kNTDW, (8) 

where W e effective noise power; 

W = actual total noise power. 

Several simple laws can be immediately deduced from equation (8). 
As the distance factor D is changed the effective power W e varies with 
the inverse square of the distance. Thus fixing W e at a point corre- 
sponding to threshold gives a variation of W with the square of the 
distance, or an increase of 6 db with every doubling of the distance. 
This law is found to hold roughly with the data presented in Figs. 5 
and 8. The simplicity of the law is lost in the more accurate formula 
of equation (5). H 

For the case of the flat noise distribution to an upper cutoff / 
equation (6) becomes: 


/. = /o/V3. (9) 


W = f W(f) . (10) 

From equation (8), 

W. = cW/f. = tf.JF(/)(V3//.) 

= crv/lFXfl, (11) 

where c = a constant, and the threshold of visibility of the noise 
is independent of the cutoff frequency f when the power per cycle, 
W(J), is kept constant. This law appeared rather startling when 
first discovered experimentally by M. W. Baldwin in tests similar 
to those plotted in Figs. 5 and 8. It means that as noise is 
added by raising the cutoff frequency of a flat distribution, the 
masking effect of the additional fine grained noise exactly com- 
pensates for the increased noise amplitude, to keep the perception 
at a constant. The law is approximately followed by the data 
for the flat distribution in Figs. 5 and 8, where it would be repre- 
sented by lines of 6 db per octave slope. The simplicity of the law is 
again lost in the formula of equation (5). The reality of the correla- 
tion which forms the basis for equation (5) is, however, shown by the 
data for the up-tilted distribution in Figs. 5 and 8. When the correla- 
tion is ignored, an equation similar to equation (11) is obtained, 
plotted with a slope of 6 db per octave in Figs. 5 and 8. The actual 
data, however, show a distinctly steeper slope, which is reasonably 
well indicated by equation (5). 

Actually the differences between the data for the flat and up-tilted 
distributions can be used to give a sensitive evaluation of the sampling 
interval DT. If Wi and TF? are respectively the total powers at 
threshold for flat and up-tilted noise up to the same cutoff frequency, 
the ratio of these, from equations (6) and (8), is: 

Wz/Wi = 3/\/5 (1.3 db). (12) 

From equation (5) the ratio is approximately: 

W z /W l T%D5P/3. (13) 

A plot of the data, equation (12), and the best fit for equation (13), 
are shown hi Fig. 9. The value of T is obtained from this best fit for 
example at the point where Wz/Wi equals 7r 2 /3 or 5.17 db. Here T 
l/(f D) = 1/4.55 = 0.22 microsecond. The plot of W 2 /W 1 from 




equation (5) without the simplifications of equation (13) is also shown 
in Fig. 9. 

It seems doubtful that the sampling area should be so sharply de- 
fined in the eye as to give a weighting function with the deep minima 
shown by the curved lines in Fig. 6. Some exploration has accordingly 
been carried out of a weighting function which starts out at low fre- 
quencies with the curves of Fig. 6, but which before the minima are 
reached translates to straight lines as shown. The differences which 


Fig. 9. Difference in perception between flat and up-tilted noise. 

result from this in the particular comparisons of Figs. 5 and 8, are 
well under 1 db. The simpler form of weighting function appears 
practical, though later checks with narrower distributions of noise 
are desirable. 

The value of DT which has been determined corresponds to 4.8 
minutes of arc at the observer's eye. The sampling area is therefore 
larger, on a side, than the conventional one minute of arc of visual 
acuity. This is not unreasonable, inasmuch as the one minute figure 
is obtained with substantially 100 per cent contrast, while here the 
sampling area merges noise grains under near threshold conditions, 


where the typical contrast is substantially less. It can be expected 
from this that the sampling area may vary in size from threshold 
conditions to random noise much above threshold. This is confirmed 
by the experimental viewing of fields of wide-band random noise. 
As the noise is reduced in amplitude to approach threshold, the char- 
acteristic granular form of the noise perceived at the higher intensities 
yields to larger floating nebulous masses. 


The threshold of perception of the difference in luminance between 
two adjacent areas is characterized, over a range, by a constant ratio 
of the difference in luminances to the greater of the two. This is 
known as the "Weber-Fechner law." 5 This threshold difference was 
further interpreted by Fechner as an elementary sensation step, thus 

dS = dB/B. (14) 

The ratio on the right-hand side is known as the "Fechner fraction." 
The expression may be integrated, as 

S = log B -f constant. (15) 

The range of validity of the Weber-Fechner law and the deviations 
from it outside this range have been the subject of much experiment. 
A general summary of some of this work has recently been presented 
by Moon and Spencer. 6 

This paper considers particularly two phases of the departures. 
The first is a gradual rise in the value of the threshold Fechner frac- 
tion toward lower luminances of the field. This is generally well 
known, and it is obvious that it must occur because at the threshold 
of perception the luminance perceived can just be distinguished from 
physical black, and the value of the Fechner fraction is therefore 1. 
Luminances just below this, hence, appear as "subjective black/' 
and the threshold luminance gives the boundary of subjective black. 

The second phase consists in the influence of glare light from a sur- 
rounding field. This is of great practical importance in daily vision 
and in the viewing of a picture, because the area being concentrated 
upon is almost never surrounded by substantial darkness. The 
treatment of this phase of the problem is much simplified by the 
"Holladay principle," which gives a weighting formula establishing 
the equivalence of any glare field distribution to a total field of con- 
stant luminance, which last is then called the "adaptation luminance." 




This formula is : 
B A = 0.923 B B + (K/Tr)fB(e, 

~ 2 cos 6 da = aB B + bB s , (16) 

where B A = adaptation luminance in millilamberts ; 
B B = luminance of test field in millilamberts; 
B s = maximum luminance existing in surround field, in milli- 

B(8, <p) = luminance of surround field over elementary solid 

angle du; 
= angle in radians between line of sight to elementary 

spot dco and line of sight to test field ; 
<p = angle measured about line of sight to test field; 
d& = elementary solid angle in steradians; 
K = a constant = 9.6 X 10~ 3 ; 
a, 6 = constants. 

The formula assumes a test field of a diameter subtending an angle 
of 1.5 degrees (.026 radians) at the eye. This is illustrated by the 
field II in Fig. 10. 

Fig. 10. Picture field. 

The first equation assigns to a in the second equation a fixed value 
0.923. The quantity b is a parameter measuring the character of the 
distribution of luminances in the surround field. It assumes its 
maximum value, namely 0.077, for a complete surround all at the lumi- 


nance of the test field B B . It approaches its minimum value of zero 
for a surround which is all physically black except for a small area of 
luminance B s , and the value further falls rapidly as this area is re- 
moved from the line of sight. Some intermediate fields are illustrated 
in Fig. 10. 

The value of b is 0.05 for a field of luminance B s entirely surround- 
ing the circle VI, the space within this, to the test field boundary, 
being physically black. Single areas of luminance B s at the spots 
III, IV, and V give values of b respectively 0.01, 0.001, and 0.0001. 
Fig. 10 also shows the outline of a picture field, viewed at D = 1 with 
aspect ratio of 3 (height) to 4 (width). 

Moon and Spencer specify the luminance of a test object enough 
lower than that of the test field to be just perceptible, and tell how 
this is influenced by the adaptation luminance. The test object sub- 
tends an angle of one degree, indicated by field I in Fig. 10. The for- 
mula breaks up into two cases according to whether the adaptation 
luminance is greater or less than the test field luminance. 

In the first case, where 

B A = aB B + bB 8 ^ B B , (17) 

the threshold is reached at a value indicated by the empirical equation 
B B - B e = A = c (A -f VaB B + bB s }\ (18) 

where B = luminance of test object in millilamberts; 
A = a constant = 0.255; 
c = a constant = 0.143. 

In the second case, where 

B A = aB B + bB 8 < B B , (19) 


A plot of the formula is illustrated in Fig. 11 in terms of the Fech- 
ner fraction (that is, AB/J5) as a function of the luminance B B of the 
test field, for a variety of values of bB s . The curves are in general 
given for integral powers of 10 for bB s , but hi one case b is given its 
maximum value (0.077) for a maximum surround luminance B s = 
10,000 millilamberts. The asymptotic value reached for bB s = is 
also illustrated. 

As in the case of the Weber-Fechner law, it is possible to integrate 
the Moon and Spencer formula to obtain the total cumulation of 
perceptible steps in luminance, say from some maximum luminance 


of the test field B B down to a lower specified value. This may be done 
by taking the approximate relationship 

dS AS_ 
dB AB 


and putting AS = 1 (to represent a single step), and by expressing 


> i- 



- .077 


'O' 4 IO' 3 IO' 2 IO' 1 


Fig. 12. Subjective Response (b = .077). 





Fig. 13. Subjective Response (6 = .01). 

A as a function of B B from equations (18) and (20). Formally, this 
leads to: 





B s = .OlmL 

1 ,,-'-' ' 












i . 




t- 1 












J _ 











>-' io' 6 to' 5 io" 4 io" 3 io" 2 io~' i 


Fig. 14. Subjective Response (6 = .001). 

In the simpler case where the Weber-Fechner law is followed, that 
is, where AB is proportional to B, the integration yields the result 
given in equation (15). 

Where the Moon and Spencer formula is assumed, the appropriate 
expression from equations (18) and (20) is introduced in the inte- 
grand. In this instance, the integration is rather laborious but 
straightforward. The results have been plotted in Figs. 12 to 15, in- 
clusive, each plot being for a different assumed value of the parameter 
b. Each plot shows the cumulative perceptible steps from the maxi- 
mum luminance value B s in the field, down to a specified ratio B B /B S 



B s = .OI mL . 







10' 3 io~ 4 io' 3 


Fig. 15. Subjective Response (6 = 0). 


26 PIERRE MERTZ January 

of this, for a variety of values of B s as indicated. The plot is contin- 
ued in each case to the approximate boundary of subjective black. 

The curves from the Moon and Spencer formula have been com- 
pared with other determinations of the deviations from the Weber- 
Fechner law. 7 The agreement is not always too satisfactory, although 
this is undoubtedly in part caused by differences in viewing condi- 
tions, which are quite varied. In a general way, one can say that in 
the present state of knowledge the form of the formula proposed is 
reasonably adequate, but the constants used in it may be subject to 
some later revision. 

The Moon and Spencer formula represents an important step for- 
ward in understanding the perception of contrast in the viewing of 
scenes and images. It, together with the Holladay principle, traces in 
a simple form the variations in this perception with picture content 
and highlight luminance level. The general qualitative facts of this 
are common knowledge, but the formula presents them in a compact 

The formula, of course, describes the perception of contrasts of 
areas of about photometric size, rather than grains of the size found 
in random noise. It has been found, however, that the vision of fine 
lines and areas is largely describable in terms of contrast perception 
in larger areas. 8 Thus the information on areas of photometric size is 
at least illustratively valid, and probably even more, in the considera- 
tion of random noise perception. 


It is clear that the susceptibility to noise in the over-all system is 
dependent only upon the characteristics of the system beyond the 
point at which the noise is introduced, up to the ultimate viewing. 
Hence the only portion of the characteristic influencing the suscepti- 
bility of the system to the noise is the transfer characteristic between 
the electrical signal at this point and the final image luminance, with 
the subjective characteristics appropriate to the image viewing condi- 
tions. This is illustrated in two forms of presentation, parts (A) and 
(B) of Fig. 16. 

In (B) the electrical signal is plotted in terms of db below the maxi- 
mum signal. In (A) the electrical signal is plotted as the arithmetical 
ratio to the maximum signal. The luminance is plotted, in both cases, 
in terms of a hypothetical photographic density by which maximum 




highlight luminance in the picture is attenuated, to equal the given 

In (A) if a noise impulse is superposed on a given signal c, it in- 
creases it to d. The increment in signal, called A 7, leads to an incre- 
ment in luminance (measured in density units) called A&. A given 
increment or decrement in voltage throughout the signal range ap- 
pears as a constant displacement in the plot, represented by the two 
dotted lines. 

In plot (B), on the other hand, a modulation of the signal which in- 
creases its level by At; (measured in decibels), raises it from / to g, 
leading to an increment in the luminance from g to h, or A6 (measured 







-2 -I 


-10 * 


1 1 



Fig. 16. Characteristic of reproducer. 

in density units). This modulation, over the signal range, appears as 
a constant increment (or decrement if in the reverse direction), and 
is again represented by two dotted lines. 

That is, additive noise leads to luminance changes (in density 

A6 = (db/dV) AF, (23) 

where AF = noise- to-signal amplitude ratio. 
Modulation leads to similar luminance changes, 

A6 = (db/dv)&v, (24) 

where Av = modulation (from undistorted signal) in decibels. 

28 PIERRE MERTZ January 

The quantities db/dV and db/dv are seen to be important in trans- 
lating the respective electrical disturbances into image disturbances. 9 
The first, namely db/dV, times a factor which will be explained, has 
been called the "interference sensitivity" of the reproducer. It is 
sometimes desirable to express equation (23) in logarithmic terms. It 

20 log 10 (A6) = 20 log 1Q (db/dV) + 20 lo glo (AF). (25) 

The last term on the right is now the noise-to-signal ratio in db. 
The term on the left can be rewritten in terms of the Fechner fraction 

A6 = 0.4343 &B/B 

20 A6 AJ5 



201ogi (A6) = 201ogi (20A&/8.686) - 20 lo glo (20/8. 686) 

= 20 logifl( AB/B) - 20 Iog 10 2.30 (26) 

Equation (25) becomes: 
20 logio(AB/B) = 20 lo& (db/dV) + 7.2 + 20 lo glo ( AF). (27) 

The first two terms on the right, taken together, have been called 
the "interference sensitivity" measured in decibels. The term on the 
left is a measure of the Fechner fraction in decilums, as used in Figs. 5 
and 8. 

In an entirely analogous way, equation (24) can also be expressed 
in logarithmic terms. It becomes: 

20 log 10 (A6) = 20 \og lo (db/dv) + 20 logio(Aw). (28) 

The last term on the right can be expressed with respect to AV/V 
as follows : 

Aw = 8.686 AF/T 

201og 10 (Az;) = 201ogio(AF/F) + 18.8. 
Hence the entire equation becomes : 
20 lo&o( AB/B) = 20 Iog 10 (e#>/^) + 26.0 + 20 log, ( AF/7). (29) 


The last term on the right now expresses how far below the signal, 
in decibels, is the modulation. The first two terms on the right, taken 
together, have been termed the "differential sensitivity" in decibels. 
The quantity on the left, as before, is a measure of the Fechner frac- 
tion in decilums. 

The interference and differential sensitivities are presented for sev- 
eral elementary picture tube characteristics in Fig. 17. For three of 
those shown, the luminance of the reproduction varies as some power 
n of the input signal voltage. For the fourth, the luminance varies 
as the exponential of the voltage. 

In order for comparisons between the four to be more meaningful, 
they have been so chosen that the signal voltage covers the same 
range, and gives the same finite luminance range (two photographic 
density units), for all the characteristics. This requires adding a 
small bias hi some cases to the voltage, otherwise to zero voltage the 
luminance would range to minus infinity. The actual equations are, 
for the variation to power n 

B = p(V + q) (30) 


q = I/ A/100 - 1; 
p = (1/1000") ; 

and for the exponential variation 

B = 10 2 V100. (31) 

The bias, q, is not shown in the right plot of Fig. 17, but it must be 
included in the computation of the db signal range in the plot on the 
left. This shows the influence of the factor p. 

For these characteristics the interference and differential sensitivi- 
ties are plotted in Fig. 18. Comment on these is reserved until their 
application, as illustrated in Fig. 19, is discussed. 

Examination of equations (27) and (29) shows that if the Fechner 
fraction, for some grade of impairment (such for example as threshold) 
for a given type of noise or modulation is known, the tolerance on 
noise-to-signal ratio (in decibels) is simply the difference between the 
Fechner fraction expressed in decilums and the interference or dif- 
ferential sensitivity. If the Weber-Fechner law holds in the region of 
interest, the Fechner fraction is a constant and the tolerance is merely 
that constant less the proper sensitivity. 








Fig. 17. Typical reproducer characteristics. 




+10 5 

-2 -I 




Fig. 18. Interference and differential sensitivities. 








o o 
ro ro 

TT 0> 






W-F LAW 1 % 









-2 -I 

Fig. 19. Noise- to-signal requirements. 

In Fig. 19 are plotted the Fechner fractions in decilums and the 
sensitivities (for n = 2) in decibels, over the picture luminance range 
which again is assumed as 100: 1, or a photographic density range of 2. 
The scales on the Fechner fraction and sensitivities are offset 30 db, as 
shown, so that the curves may fall in approximately the same part of 
the plot. The Fechner fractions are shown at 2 per cent and 1 per 

32 PIERRE MERTZ January- 

cent, assuming constancy in each case, following the Weber-Fechner 

For such conditions, the additive noise will be visible only in the 
deep blacks (b = -2), and will have to be at 30 + 29.1 db = 59.1 
db below the signal to be just visible with a Fechner fraction of 2 per 
cent. Modulation will be equally visible throughout the range of pic- 
ture luminances, and will have to be at 30 -+- 10 = 40 db below the 
signal to be just visible with a Fechner fraction of 2 per cent. 

Instead of the Weber-Fechner law the Moon and Spencer formula 
may be used, with the reservations which have already been noted. 
A plot is shown in Fig. 19, copied from Fig. 11 The parameter b of 
equation (16) is taken as equal to .01 and the highlight luminance B s 
as 100 millilamberts. This gives bB s = 1, and corresponds to a rather 
brightish picture. If the curve is assumed applicable to the noise 
threshold under consideration it indicates that while the noise will be 
most visible only in the deep blacks, it will be nearly as visible over 
about half the density range of the picture. At threshold the noise will 
be 30 + 24.7 = 54.7 db below the signal. 

The modulation in this case will be about equally visible over the 
upper half of the density range of the picture, and at threshold will be 
30 + 13.2 = 43.2 db below the signal. 

With this illustration in mind it is now possible to make some gen- 
eral observations regarding Figs. 18 and 19. 

1. With the assumption of the Weber-Fechner law, additive noise 
is generally most visible in the extreme blacks. With the exponen- 
tial characteristic a limit characteristic is reached in which the noise is 
equally visible over the entire tone range of the picture. With devi- 
ations from the Weber-Fechner law, the noise susceptibility tends to 
be less sharply localized, and the maximum is apt to be shifted to the 
dark grays. In the limit of the exponential characteristic the maxi- 
mum is broad and shifted to the white regions. 

2. With the assumption of the Weber-Fechner law, modulation is 
generally equally visible over the entire tone range of the picture, al- 
though in the limiting exponential case the visibility is greatest in the 
extreme whites. With deviations from the Weber-Fechner law, modu- 
lation generally becomes visible over broad white regions or the ex- 
treme whites. 

3. The characteristic f or n = 1 gives, of those considered, the 
greatest susceptibility to additive noise, and the least to modulation. 

4. As n is increased, the susceptibility to additive noise is reduced 


while that to modulation is increased. The changes are slow beyond 
n = 2. The two susceptibilities become approximately comparable in 
the limiting exponential case. 

A diagram of the type of Fig. 19 is applicable for each case of the 
data of Figs. 5 and 8. Those particular data were taken with a flat 
field, hence each point corresponds roughly with the intersection of an 
appropriate Fechner fraction curve (either following the Weber- 
Fechner law or the Moon and Spencer formula) with the vertical axis 
at highlight luminance value. From the vertical distance to the ap- 
propriate picture tube characteristic it is possible to translate the or- 
dinates of Figs. 5 and 8 to electrical signal to noise ratios for the nea,r 
threshold condition. It is further necessary for setting tolerances on 
the noise, to determine the vertical distance between the two chosen 
curves, along the ordinate to the left of maximum highlight luminance 
corresponding to the maximum susceptibility to the noise. 

The number and types of characteristics considered here are of 
course extremely limited. There are few receiving mechanisms that 
follow a pure law as assumed. In addition, negative modulation 
characteristics represent special problems. A more complete study of 
the subject should also include an examination of the point that, as the 
density range over which a given noise is rendered susceptible is in- 
creased, its general frequency of occurrence, and the geometrical area 
over which it is visible on any one picture, are both increased. 


(1) O. H. Schade, "Electro-optical characteristics of television," RCA Rev., vol. 
9, Parts I, II, III, and IV, pp. 5, 245,590, and 653; 1948. 

A. Rose, "The sensitivity performance of the human eye on an absolute 
scale," /. Opt. Soc. Amer., vol. 38, p. 196; February, 1948; Jour. SMPE, vol. 47, 
p. 273; October, 1946. 

R. B. Janes, R. E. Johnson, and R. S. Moore, "Development and perfor- 
mance of television camera tubes," RCA Rev., vol. 10, p. 191; June, 1949. 

(2) C. E. K. Mees, The Theory of the Photographic Process, Macmillan, New 
York, 1942. 

L. A. Jones and G. C. Higgins, "Photographic granularity and graininess," 
I. "The relationship between the granularity and graininess of de- 
veloped photographic materials," J. Opt. Soc. Amer., vol. 35, p. 435; 1945. 

II. "The effects of variations in instrumental and analytical techniques," 
/. Opt. Soc. Amer., vol. 36, p. 203; 1946. 

III. "Some characteristics of the visual system of importance in the evalu- 
ation of graininess and granularity," /. Opt. Soc. Amer., vol. 37, p. 217; 1947. 

IV. "Visual acuity thresholds; dynamic versus static assumptions," 
J. Opt. Soc. Amer., vol. 38, p. 398; 1948. 


(3) S. O. Rice, "Mathematical analysis of random noise," Bell Sys. Tech. J., 
vol. 23, p. 282; July, 1944; vol. 24, p. 46; January, 1945. 

(4) P. Mertz and F. Gray, "A theory of scanning," Belt Sys. Tech. J., vol. 13, p. 
464; July, 1934. # 

S. O. Rice, "Filtered thermal noise, fluctuation of energy as a function of in- 
terval length," /. Acous. Soc. Amer., vol. 14, Part 1, p. 216; April, 1943. 

(5) E. H. Weber, Archiv Anat. u. Physiol, 1835, p. 152; "Der Tastsinn u. d. 
Gemeingefiihl," in Wagner's Handworterbuch d. Physiol., vol. 3, Pt. 2, 1846, p. 
481 (Braunschweig, Berlin). 

G. T. Fechner, "Uber ein psychophysiches Grundgesetz," in K. Sachs, 
Ges. d. Wiss. zu Leipzig, Abb. 4, 1859, p. 457; Elements der Psychophysik, 1860 

(6) P. Moon and D. E. Spencer, "The visual effect of nonuniform surrounds," 
/. Opt. Soc. Amer., vol. 35, p. 233; March, 1945. 

(7) F. G. H. Pitt, Proceedings of Physical Society of London, vol. 51, p. 817; 
September 1, 1939. 

P. G. Nutting, Scientific Publications from the Kodak Research Laboratories, 
vol. 2, p. 94; 1916. 

. E. M. Lowry, Scientific Publications from the Kodak Research Laboratories, 

vol. 13, p. 138; 1929. 

H. Frieser and W. Munch, Kinotechnik, vol. 20, p. 85; April, 1938. 

(8) S. Hecht, S. Ross, and C. G. Mueller, "Visibility of lines and squares at 
high brightness," /. Opt. Soc. Amer:, vol. 37, p. 500; 1947. 

G. M. Byram, "The physiological and photochemical basis of visual re- 
solving power," Parts I and II, /. Opt. Soc. Amer., vol. 34, pp. 571 and 714; 1944. 

S. Hecht and E. U. Mintz, "The visibility of single lines at various illumina- 
tions, and the retinal basis of visual resolution," J. Gen. Phys., vol. 22, p. 593; 1939. 

(9) An independent development of equations (23) and (24) has been infor- 
mally presented by B. M. Oliver to the Television Systems Committee of RMA. 

An Improved Photomultiplier Tube 
Color Densitometer 



Summary Previously, attempts were made to modify black-and-white den- 
sitometers to make them suitable for color measurements, but considerable 
difficulty was experienced with such modifications. Therefore, the Ansco 
Laboratories developed new electronic circuits utilizing the full capabilities 
of the photomultiplier tube for this purpose. These developments, combined 
with other refinements, have made it possible to design a densitometer 
capable of using sharp-cutting filters to measure color densities up to 4.0 
and over. 


UNTIL quite recently, the photographic industry has been concerned 
mainly with black-and-white reproduction. This situation is 
reflected in the attention devoted to black-and-white sensitometry. 
However, during the past few years, the importance of color products 
has risen rapidly and in the case of motion picture film there are 
numerous commercial color processes now on the market and the 
number of production releases on color is growing continuously. As 
a result of this trend there is a corresponding demand for appropriate 
testing and control instruments and techniques. 

One important aspect of three-layer color-film sensitometry is the 
measurement of color densities of the processed test strips. In sharp 
contrast to black-and-white densitometry, wherein it is permissible 
to use the entire radiation from a given light source to excite 
the phototube receiver, in color densitometry it is desirable to make 
density measurements of the materials using narrow spectral-energy 
bands, preferably at single wavelengths. This requirement rules out 
the possibility of using relatively simple receiving systems of limited 
sensitivity or relatively inefficient optical systems a luxury available 
only in black-and-white densitometery. When one confines the 
radiation to a desirably narrow band, he is confronted with a reduc- 
tion of the available energy by a factor of 100 to 1000 or more before 
absorption by the specimen itself is even considered. Therefore, 

PRESENTED: October 28, 1948, at the SMPE Convention in Washington. 


36 MONROE H. SWEET January 

relatively drastic changes are necessary in the design of instruments 
which are required to do an equally good job in color as had previously 
been done in black and white. 


In order to meet the problem of color densitometry, Evans proposed 
the use of a cleverly modified visual densitometer and reported it to 
this group. 1 Later, null techniques incorporating high-gain ampli- 
fiers of special design were used in some objective instruments to 
achieve the necessary sensitivity. Instruments of this type were too 
slow to use for routine work although they were satisfactory for some 
applications where the volume of work was relatively small. In order 
to overcome their disadvantages, at Ansco a direct reading black- 
and-white densitometer was modified in such a way as to provide the 
necessary high sensitivity and yet preserve its other major features. 2 

Fig. 1 Explanatory diagram of elec- 
tron-multiplier-tube operation. 


In this instrument an electron-multiplier type of phototube was 
substituted for the simple phototube used in the original. This new 
tube operates according to the principles shown in Fig. 1. For each 
electron emitted by the action of light on the cathode, on the average 
two or more electrons are emitted by secondary emission from each 
succeeding dynode element. Therefore, the inpiut current is amplified 
many times when it reaches the anode. The use of the photomul- 
tiplier tube increased the sensitivity of the original instrument by a 
factor of 10,000 but it was necessary to take special precautions to 
shield and insulate the entire multiplier-tube supply and the multi- 
plier tube itself because of unfavorable polarity relationships. In 
spite of a consequent tendency toward instability the instrument was 
put into routine use and it quickly demonstrated the value of a simple 
direct-reading densitometer in color development and color-control 


In the course of further development work, a photomultiplier-tube 
circuit was devised which retained the high sensitivity of the modified 
instrument described above and in which the inherent stability was 
greatly improved. A compensating circuit was also developed which 
enables the user to calibrate the scale of the instrument to agree closely 
with virtually any desired reference standard. 


Photomultiplier tubes have acquired a reputation in some circles 
for instability. This is because many mutliplier tubes, when operated 
at a constant dynode voltage, show pronounced fatigue at even moder- 
ate light levels. 3 If one attempts to cover a density range of to 3 
by conventional use of the photomultiplier tube he often encounters 
fatigue at the high light levels (low densities) and dark current at the 
lowest light levels (high densities). It is known that the fatigue 
effects are most serious in the last few dynode stages and occur when- 
ever the current is of large magnitude and changes appreciably. In 
the present case the anode is operated at a constant-current level and. the 
last few dynodes are operated at comparatively constant currents regardless 
of the densities being measured. Therefore, the stability of the circuit 
is comparable to that of a multiplier tube operated in conventional 
circuits, but maintained under ideal optical and electrical conditions; 
namely, at constant dynode voltage and with a constant level of 
incident flux. 

Furthermore, when operated in conventional circuits, photomulti- 
plier tubes require a precision-stabilized high-voltage source. The 
present circuit completely avoids the necessity for such a source. 


The basic operating principles of the photomultiplier-tube circuit 
can best be demonstrated by reference to Fig. 2. In this illustrative 
circuit the operator manually adjusts the voltage applied to the multi- 
plier dynodes in such a way as to keep the multiplier-tube output 
current constant at all light levels. When a given specimen density 
is inserted in the light beam, the multiplier-tube output current is at 
first reduced but is then restored to its original value by increasing 
the dynode voltage. A voltmeter which responds to the dynode 
voltage applied to the tube can be calibrated in terms of density and 

* Protected by United States Patents 2,478,163 and 2,457,747. Patents on 
other novel features are pending. 



Fig. 2 Illustrative dynode-voltage-feedback circuit. 


Fig. 3 Relationship between dynode 
voltage and log (sensitivity) for a typi- 
cal photomultiplier tube. 

the scale will be fairly uniform 
since the relationship between 
phototube sensitivity and dynode 
voltage is virtually logarithmicf 
as shown in Fig. 3. In actual 
practice an electronic tube per- 
forms the dynode voltage adjust- 
ment automatically and instan- 
taneously. Therefore, in effect 
the sensitivity of the multiplier 
tube is continuously adjusted so 
that when the light intensity is 
high the gain of the tube is low 
and vice versa. The product of 
light intensity and tube sensi- 
tivity is at all times constant. 

Fig. 4 is a simplified schematic 
diagram of the densitometer cir- 
cuit. The voltage applied to the 

931-A photomultiplier-tube dynodes is derived from the drop across 

t Density, D = logio (l/F) = logio (0). In which unit flux is incident on the 
specimen, F is the flux transmitted, and is the opacity of the specimen. 




the cathode resistor R of the type 807 control tube. This voltage is 
controlled by the grid G whose potential is determined by the anode 
current of the photomultiplier tube (by virtue of the voltage drop 
it creates across the grid resistor R'). 

The electrical-polarity relationships are such that, as illustrated, 
an increase in illumination on the phototube causes the voltage across 
R to drop and therefore the effective sensitivity of the 931-A tube to 
decrease. This negative-feedback action is continuous and therefore 
the voltage developed across R is a reliable measure of the phototube 

Fig. 4 Simplified schematic diagram of multiplier-tube feedback 

illumination. A few of the specific advantages obtained from the use 
of this circuit are : 1. High sensitivity with high stability. 2. An 
approximately logarithmic electrical response over a wide range of 
light levels. 3. Reduction of photomultiplier-response fatigue. 
4. Elimination of the need for a stabilized high-voltage power 

Dynode Voltage Versus Density Relationship 

From the qualitative description of the instrument given above it 
will be recognized that the uncompensated output of the instrument 

40 MONROE H. SWEET January 

portrays the relationship between dynode voltage and photomulti- 
plier-tube sensitivity. The equation, 

S = k-E/ 2 (1) 

where S = net sensitivity of tube in terms of anode current per unit 

incident radiant flux 

k = a constant, characteristic of the tube 
E = dynode voltage 
n number of stages 

is often used to express this function* but it is not immediately appar- 
ent that E versus S is even quasi logarithmic, 

since logio S (n/2) logio E + log k. 

However, the following treatment shows that when n/2 is large, 
the relationship between E and log S approaches linearity over the 
finite range of operating values of S encountered in practice. 

In the present discussion it is assumed that the anode current is 
held constant in the presence of variations in incident flux on the 
photomultiplier tube by control of the dynode voltage. 


F-S = KorS = K/F (2) 

where F is the radiant flux received by the photosurface and K is a 

By definition, the optical density of the specimen is 

D =* logio (1/F) 

where unit flux is incident on the specimen and F represents the trans- 
mitted flux received by the phototube and from (2) in the present 

log S = log K + (log (1/F) = D) 
from (1) 

log S = log k + n/2 log E 

D = n/2 log E + K 1 (4) 


K f = log k - log K 

* C. C. Larson and H. Salinger, "photo-cell multiplier tubes," Rev. Sci. Instr., 
vol. 11, pp. 227; July, 1940. The fundamental photomultiplier tube equation is 
S = k X G n , in which G is the gain per stage. 



dD/dE = (n/2) (lo glo e) (1/E). (5) 

The ratio of the slope (dD/dE) taken at a specimen density of 2.0 
to that taken at density 0.0 is a convenient measure of the "linearity" 
of the system. Letting M represent this ratio, 

M = --; from (1) B = 


therefore M - K"/2)(loB* )(fc/S)/]. .. /S,oV' 
[(n/2)(lo glo " " 

as the number of dynode stages n increases, M ap- 
proaches unity as a limit and the voltage is linear with density. 

In the 931-A tube 2/n ^ 0.23 and if D = and Z) 2 = 2.0, (&/&) 
= 1/100. Under these circumstances M = 0.36, whereas for a con- 
ventional circuit operated at constant dynode voltage M = 0.01 for 
the same density range. 

The compensating circuit described in a subsequent section pro- 
vides the correction necessary to maintain a constant slope of the 
dynode voltage versus density curve over the operating range of the 


It was stated earlier that the performance of the instrument is 
unaffected by ordinary variations in the high-voltage power supply. 
This fact will become evident from the following proof of independ- 
ence of the instrument's performance with respect to changes in the 
amplifying characteristics of the control tube. 

In Fig. 5* the photomultiplier tube is shown as P. E represents 
the voltage developed across the dynode resistors of resistance R. I 
is the 807 control tube plate-cathode current. E b is the voltage de- 
veloped across the gaseous stabilizer tube of the actual circuit and is 
here represented as a battery to provide a suitable operating poten- 
tial between the photomultiplier-tube anode and dynode No. 9. 
E c is the voltage applied to the photomultiplier-tube load resistor 
r and is referred to ground. The photomultiplier anode current is 

* In the actual circuit, a cathode-follower tube is inserted between point X 
and the 807 grid but it has no significant effect on the analytic behavior of the 
basic circuit treated here. 




Fig. 5 Explanatory diagram which illustrates the analytical 
behavior of the dynode-voltage-feedback circuit. 

i and F is the flux incident on the photo cathode. As drawn, the 
following relationships apply : 

rs, = E n -E t -i 

E,, = E c - i-r 
IE, = E c - E b - i-r = E, - i-r < Oj 



E a = E e E b . 

Since the sensitivity, S, of the photomultiplier tube may be defined as 

S = t/F, then (from (1)) i = F-k-E/ 2 

E = I.R = (g m -E )-R (2) 



E = g m -R[E. - F-k-r-E"' 2 ] 



= -fc-r-Jg*/* (g) 

(1/JB-flfJ + n/Z-F-k-r-E"/ 2 - 1 ' 

Equation (9) shows that if (R-g m ) is large, the relationship be- 
tween E and F will be independent of R and g m . Furthermore, since 
variations in the high-voltage supply to the 807 control tube can 
effect the circuit only by changing the effective value of g m , the 
present circuit likewise is independent of fluctuations in the 807 
plate-supply voltage. 


As mentioned in the introduction, a convenient optical system of 
arbitrary geometry and relatively high efficiency was adopted but the 
resulting density values obtained, particularly with scattering speci- 
mens, do not perfectly conform with those obtained according to 
techniques prescribed by the American Standards Association for 
the determination of Diffuse Density. 4 For this reason alone it would 
be desirable to provide some automatic scale compensating feature. 
In addition, the fact that the relationship between E and (log F) is 
not quite linear in the case of the 931-A makes such a feature even 
more desirable since it can be used to correct both distortions simul- 

Therefore, a circuit was developed which corrects the output 
voltage in such a way as to give a close approximation to the desired 
calibration ( 0.02 over the density range 0.0 to 3.0). Its basis of 
operation is illustrated in Fig. 6. 

The output meter measures the total dynode voltage AB with the 
indicated electrical polarity so that Z is always negative with respect 
to A and the magnitude of the voltage E AZ is of course directly 
proportional to the current flowing through R 2 . 

Now if there were no compensation, the family of possible speci- 
men-density versus meter-reading curves which could be obtained 
by control of shunt S is shown in Fig. 7, where curve A represents the 
lowest shunt resistance and curve F the highest. The heavy straight 
line represents the ideal relation. It will be noted that curve E 
gives reasonable agreement with the ideal values over the density 
range to 0.5. 

If it were possible in the case of curve E to readjust continuously 





oc sou/rce 

Fig. 6 Circuit which provides corrective action resulting in a uni- 
form relation between specimen density and meter current. 


.6 .8 2.0 



Fig. 8 Comparison of relationship 
between specimen density and result- 
Fig. 7 Family of specimen-density versus ing meter current after compensation, 
meter-current curves obtained by varying shunt (Errors exaggerated for illustrative 
resistor S of Fig. 6. purposes.) 

the shunt S to a lower resistance value at all points corresponding to 
density 0.5 and above, it is clear that the approach to the ideal rela- 
tion could be extended over a wider density range. 

This is done, in effect, by the action of the compensating circuit 
shown in Fig. 6. Voltage E A z rises continuously with increasing 
density. Voltage VCD is stabilized by the voltage-regulator tube and 
resistor R* is a relatively low impedance unit. If the variable tap 


of Rs is so adjusted that voltage E AZ = voltage E AX when the specimen 
density is 1.0 then at all higher densities (and therefore higher E AZ 
voltages) rectifier M will pass current because the voltage difference 
EXZ will be of proper polarity for conduction. Furthermore, if 
rectifier M is of low resistance it may be regarded as a switch which 
closes whenever voltage E AZ exceeds E AX and opens when E AZ is less 
than E AX . Under these conditions whenever the specimen density 
rises above 1.0, Rs will act as a shunt path for the meter and if 
its value is properly selected a curve of the type shown in Fig. 8 be- 
tween and 2 will result. The range of satisfactory calibration may 
thereby be extended to cover the range 0-2. The correction is auto- 
matic, reliable, and instantaneous. 

By adjusting the variable tap of R so that voltage E AZ =E AY 
when the specimen density is 2 the correction can be extended up to 
density 3. The number of corrective steps that could be used ob- 
viously is unlimited. However, two such shunts perform, the cor- 
rection satisfactorily. 

Thus the compensating circuit provides an output current that is 
linear with density, making the instrument uniquely suited for 
automatic linear density recording by connecting the output to any 
standard ink recording milliammeter. 

The circuit used in the actual instrument incorporates two inde- 
pendent sets of compensating circuit elements. One set is used when 
reading the densities of ordinary silver images and is adjusted at the 
factory so as to give results which are in approximate agreement with 
those obtained by ASA Diffuse-Printing-Density Type P-2b.* The 
second set is adjusted so as to give results which are in agreement with 
those obtained by a proposal submitted to the ASA for color densitom- 
etry.f A switch located on the right-hand side of the case permits 

* In spite of the fact that the geometry of the optical system does not conform 
with that under which primary diffuse printing density measurements of the 
American Standards Association are made, the agreement has been shown to be 
sufficiently good for samples of widely different grain size as to permit the small 
errors to be ignored in routine sensitometric work. Specific data concerning 
the differences resulting from the use of an instrument incorporating a similar 
optical system have been reported earlier. 5 

t The order of the agreement obtained in this way is considered satisfactory for 
routine photographic sensitometry. Departures from the ideal values are kept 
at a minimum by virtue of the fact that the dyes used in most commercial proc- 
esses are not sufficiently sharp cutting to cause serious errors in the results in 
the present case where the spectral purity of the source-filter-receiver combination 
is relatively high. 

46 MONKOE H. SWEET January 

the selection of either circuit. The principal reason which necessi- 
tates using two circuits is the difference in diffusion between the silver 
and color film specimens; i.e., silver images have a relatively high 
scattering power whereas ordinary color-film samples scatter little 
light and the corresponding differences in effective density versus 
standard density require different degrees of compensation. Either 
one or both sets of controls can be readjusted, in a matter of minutes, 
to bring a given instrument into agreement with values obtained by 
the use of some other standard. 

In its commercial form the bias points X and Y are both adjustable 
by potentiometers as are also resistors R$ and #4. This involves eight 
controls in all, which are accessible through a door on the right-hand 
side of the instrument case. 

When used as a photometer for measuring external light levels or 
as a reflection densitometer, it is desirable to have the meter respond 
uniformly to uniform changes in log (incident flux). This require- 
ment is well satisfied when the compensating circuit switch is placed 
in the "color" position. 


' 'Fatigue" is an undesirable characteristic of many electronic proc- 
esses. As an example, when one irradiates a barrier-layer photocell 
with bright light the initial output current may be relatively high, 
but the response usually will fall off (exponentially) with time until 
eventually it reaches a stable value. 

In the case of the electron-multiplier phototube similar effects 
may be found which are attributable to the fatigue of the secondary 
emissive dynodes. Although the photosurface itself may contribute 
to the over-all fatigue effect too, this is not the usual case in photo- 
multiplier-tube operation because the level of incident radiant energy 
is very low. 

Qualitivatively, the benefits of the feedback circuit, as a means for 
reducing fatigue, may be appreciated by the following argument : 

1. In practice, virtually any electronic device such as a multiplier 
tube will provide a stable output after a given time if all pertinent 
operating conditions are held constant. In the present case if the 
dynode voltage is stabilized and the incident flux on the photosurface 
is held constant, after a certain period the anode-current output will 
reach an equilibrium value. 

2. Fatigue effects (which in photomultiplier tubes are confined to 


the dynodes) will be some function of recent changes in electron bom- 
bardment of each of the dynodes involved. This is of course equiv- 
alent to a statement that the yield in secondary-emission ratio de- 
pends on the immediate history of the incident bombardment current 
on the dynode surface. It can therefore be appreciated that the 
magnitude of the fatigue in any specific case is a function of the mag- 
nitude of the disturbance from previous equilibrium conditions. 

3. In the conventional or constant dynode voltage operation each 
dynode experiences major changes in bombardment current which are 
directly proportional to the changes in the flux level incident on a 
phototube. In the case of inverse-feedback operation the change in 
initial bombardment current for any given dynode, for an identical 
change in incident radiation on the photosurface, is less than in the 
constant dynode voltage operation case because the dynode voltage 
is always simultaneously changed in such a direction as to tend to 
maintain the dynode currents constant. 

4. If now, we plot, for the inverse-feedback operation case a curve 
which relates dynode voltage to incident flux and consider any par- 
ticular point on the curve, an arbitrary displacement along the 
curve will correspond to a specific change in flux level and a specific 
change in dynode voltage. Or conversely, if as the result of fatigue 
effects the tube sensitivity is reduced to a fixed amount, the change 
in dynode voltage necessary to restore the original output can be de- 
termined readily. Therefore, it is convenient to think of the changes 
in voltage in the inverse-feedback operation case in terms of the 
equivalent changes in incident optical flux. 

5. Since in every case where comparable conditions of incident 
flux exist inverse-feedback operation will result in smaller correspond- 
ing changes in dynode bombardment current than in the constant 
dynode voltage operation case, the consequent fatigue effects, (as 
measured in terms of the increase in flux level necessary to restore 
the initial anode current in the case of constant dynode voltage opera- 
tion and the initial dynode voltage in the case of inverse-feedback 
operation), will also be less. 

The above analysis demonstrates the superiority of inverse-feed- 
back operation qualitatively. In the following quantitative analy- 
sis, it will be shown mathematically that the difference between 
dynode bombardment currents in the case of constant dynode voltage 
operation and inverse-feedback operation is negligible for the first 
five or six stages but that in subsequent stages the difference becomes 

48 MONROE H. SWEET January 

significant. It is assumed throughout this discussion that in the 
case of inverse-feedback the amplification of the control tube is 
infinite and therefore that the photomultiplier-tube output current is 
constant regardless of the incident-flux level. 

The analytical expressions for the anode current, as a function of 
fatigue, are as follows: 

For either constant dynode voltage or inverse-feedback operation 

- k-q f . c .(l -- a-')] (?[!- 

[1 - kg(l - a-')]} 

= I P . C . [0(2 - ft)- 0(1 - ft) . . . 0(1 - ft)] 

I t = anode current at tune t 

Ip.c. = photosurface current produced by incident flux 

ft', Gz ... Gg r = gain of each dynode after elapsed time t 

G = gain per stage before onset of fatigue and is assumed to be 

the same, initially, for all dynodes 
Q "fatigue factor" and in this case is chosen equal to k-q- (1 - 

a~') where 
k a constant which determines the upper limit (G n f ) to which 

dynode stage n will fatigue when t = < 
q = a function of the current bombarding the particular dynode 

in question and therefore depends on the product of the 

preceding terms of the equation. 
In this case it is chosen as equal to \i n where i n represents 

the bombarding current for the nth dynode. 
a = a constant which determines the rate of fatigue with time 
t the time elapsed since the onset of fatigue 
The ratio of the initial to final outputs is 

.. ...- kq B ) 


Let R' be a measure of the decrease in fatigue and equal 
RCDO [(1 - fcgi)(l - kqz) ... (1 - kq s )] CDO 

R IFO [(I - 

where RCDO = the ratio R, above, for the case of constant dynode 
voltage operation and RIFO is for the inverse feedback case. 

Now if k = 0.1 and j8 is taken as [0.1( Az'A')(l a-')]* and a change 
of illumination of 1 : 1000 is considered, then 

p, (l-O-l) 9 _ .nog 

' (1 - 0.1) 6 (0.91)(0.92)(0.95) 


The treatment is considerably simplified if we allow the time t 
to approach infinity and confine our attention to initial and final 
values of output which result from a given change in flux level. It is 
also convenient to ignore the small change in ft which takes place as a 
consequence of the slight reduction in actual bombardment currents 
during the fatigue process itself. With these simplifications, for a 
change in incident light of 1:1000, in the constant dynode voltage 
operation case the output will fatigue to 0.37 of its initial value whereas 
in the inverse-feedback operation case the output will fatigue to 
only 0.43 of its initial value. Therefore, operation of the tube in 
an inverse-feedback circuit represents a definite improvement 
with respect to fatigue of 1:0.88. It should be noted that in the 
treatment of the inverse-feedback operation, the anode current 
output is investigated as though no feedback were present, once the 
total level of the incident flux has changed to its new level and the 
general magnitude of the corresponding dynode-amplification factors 
have been established. 

From the above discussions it is clear that feedback operation 
presents a real advantage from the standpoint of fatigue at all light 
levels although its advantage in this respect is greatest when ex- 
tremely large differences in operating flux levels are involved. 


Fig. 9 shows the actual circuit. The high-voltage power supply 
for the multiplier tube is conventional but it is not stabilized since 
the associated control circuit automatically compensates for line- 
voltage fluctuation. A Type 807 beam-power pentode serves as the 
control tube for the dynode voltage. Its plate-cathode circuit in- 
cludes the dynode-voltage dropping resistors which divide the total 
dynode voltage into equal increments for distribution to the indi- 
vidual dynodes. A 12SF5 triode is inserted between the photomulti- 
plier-tube anode circuit and the 807 control grid and acts as a cathode 
follower. The operating conditions of the 12SF5 are such that its 
grid current is at all times less than 10~ 8 ampere yet its output 
impedance is low enough that the development of slight grid current 
during the life of the 807 control tube will not disturb the performance 
of the instrument as a whole. 

* Results obtained for R' with this value ofV are very nearly the same as those 
obtained with = k (Az)(l a~ r ) and = k(id)(\ a~ l ) since the /3 values 
for the first 6 stages are nearly identical. The electron bombardment of the 
first dynode stage is identical for both inverse-feedback and constant dynode 
voltage operation at equal light levels. 





Experience shows it is desirable to operate the multiplier tube with 
a constant dynode No. 9-to-anode voltage of between 50 and 80 
volts. A constant-voltage gas tube, inserted between the 807 cathode 
and dynode No. 9 terminal, performs this function satisfactorily. 
Over the entire operating range of the device the voltage swing of the 
photomultiplier tube anode-to-807 cathode is less than 5 volts. Since 
the 807 cathode-dynode No. 9 voltage is sensibly constant the photo- 
multiplier tube anode-to-dynode No. 9 voltage swing is only slightly 
more than 5 volts. 

If the voltage applied to the photomultiplier anode load resistor 
is sufficiently high the anode current will automatically be held nearly 
constant throughout the operating range of the instrument. The 
voltage actually applied is more than 100 volts and therefore the 
variation in anode current is of the order of only 5 per cent. 

It is necessary to provide a "bucking current" to counteract the 
current developed in the dynode- voltage-measuring circuit when there 
is maximum incident flux or zero specimen density (minimum dynode 
voltage). This is conveniently obtained from the stabilized 807 
screen-voltage supply. The bucking current is approximately 2 
milliamperes and the normal range of applied dynode voltage is 30 
to 90 volts per stage. 

The theory of the compensating circuit has already been discussed. 
A type 6H6 twin-diode vacuum rectifier serves its purpose in this 
circuit very well since it has relatively low forward resistance and 
passes negligible reverse current. All resistors except the photo- 
multiplier anode load resistor are wire-wound. 

To provide power for various attachments, stabilized low voltage 
is accessible through a door located at binding posts at the rear of the 

Depending upon the sensitivity of the individual instrument, the 
lamp is connected to its source through different taps of a resistor. 
In all cases, stepless control of lamp brightness is affected by opera- 
tion of a gear-driven solenoid which serves as a "zero" control. The 
output meter is a standard Weston Model 273 fan-shaped milliam- 
meter having a long scale and high speed of response. 

The basic instrument is believed to provide the highest sensitivity 
of any general purpose commercial photometer,* and there are many 
desirable ways in which this feature can be used. In the present case, 

* The average Model 12 densitometer reads full scale (density 3.0) with an 
excitation of only 0.1 microlumen of energy 2870 degrees Kelvin. 






this sensitivity is used principally to incorporate sharp cutting, but 
very dense, gelatin-foil niters in the path of the light beam in order to 
obtain high spectral purity without resort to monochromators, inter- 
ference filters, gaseous discharge sources, or the like. 


With reference to Fig. 10, light from a 6-volt concentrated filament 
automobile lamp, controlled as indicated in the preceding section, is 
collimated by an aspheric lens after passing through a glass infrared 
absorbing filter. The beam then passes through la liquid cupric 
chloride filter which absorbs the radiation of wavelengths greater than 

645 millimicrons but transmits 
virtually all of the shorter wave- 
length radiation. A second con- 
denser focuses the beam on a 3- 
mm diameter aperture mounted 
in the top plate of the instrument 
proper. After partial absorption 
by the specimen the light is fur- 
ther absorbed by the "color fil- 
ter" and is finally intercepted by 
the photomultiplier-tube cathode 
surface. The "color filter" actually 
consists of a pack of several small 
gelatin-foil filters which serve to 
confine the continuous spectrum 
of energy emitted from the tung- 
sten source to each of the desired 
wavelengths. There are six sets 
of filters, each set being located 
over a different aperture in a 
"filter cylinder" surrounding the phototube. 

In one position a red-filter pack is used which serves to absorb 
radiation of wavelengths shorter than 644 millimicrons. In this 
case, the cupric chloride filter serves as the long-wave absorber with 
the result that the radiation reaching the photosurface is nearly pure 
spectral energy of 644 millimicrons wavelength. 

In another position the filter pack is such as to confine the trans- 
mitted radiation to spectral energy of wavelength 546 millimicrons. 
Similarly in a third position the transmitted wavelength is 436 milli- 



10 Optical system of Ansco 
Color densitometer. 


In the fourth position a "Visual" filter is interposed. This filter 
is of such design as to reproduce the response of the eye when taken 
in combination with the spectral characteristics of the cupric chloride 
liquid filter, and the spectral sensitivity of the photomultiplier tube. 

A fifth filter position labeled "3" interposes enough neutral density 
that the flux received by the phototube comes within the range of 
operation of the instrument in view of the finite range of adjustment 
of the lamp-intensity control. This permits black-and-white den- 
sities from to 3 to be measured directly. The sixth filter position, 
labeled 6, uncovers the phototube completely and in this case when a 
specimen density of 3.0 is placed in the sample position the meter 
reading can be brought to zero thereby permitting black-and-white 
densities from 3 to 6 to be measured. Ordinarily, there is sufficient 
latitude of flux control to permit measurement of densities up to 

The spectral characteristics of the combined source-phototube- 
filter-receiver products for the different filter cylinder positions of 
recent production units are illustrated in Figs. 11 and 12. It is 
believed that the visual filter combination shown in Fig. 12 represents 
one of the best approximations to the response of the eye yet obtained 
with a phototube having a spectral response similar to the average 
S-4 surface 

Although at the present time there is no standard for color densi- 
tometry (either ASA or SMPE) nor even a recommended practice, 
a proposal has been made to the American Standards Association for 
their consideration for adoption as an American Standard. This 
proposal specifies that for the densitometry of three-layer monopack 
color film the measurements shall be made at the three wavelengths 
corresponding to the prominent mercury and cadmium lines of 
emission which fall at 436, 546, and 644 millimicrons. These wave- 
lengths lie close to the spectral density peaks of the average commercial 
color film. Furthermore, there is a good reason to believe that any 
satisfactory three-color process would of necessity have absorption 
peaks falling close to those specified. Therefore, it was a design 
goal to provide narrow band isolation filters which coincided in peak 
transmission with the proposed standard. 

In Figures 11 and 12 the maximum log reciprocal (relative 
response) for each of the three blue, green, and red filters has been 
adjusted to zero. Actually, the minimum filter density, in the case of 
the blue filter, is approximately 4.0 at 436 millimicrons and since this 




is virtually monochromatic, less than 1 part of 10,000,000 of the total 
energy of the initial beam is ultimately received by the phototube. 
It is somewhat difficult to design a satisfactory red-filter combination 
in spite of the greater relative energy emission of the tungsten lamp, 

4OO 2O 4O 6O SO &O 2O 4O 6O 8O 6OO 2O 4O 6O 8O 7OO 


Fig. 11 Log (reciprocal relative response) versus wave- 
length for the three-color filters used for transmission meas- 
urements in the Ansco Color densitometer. The wave- 
lengths corresponding to peak transmission coincide with 
those of a proposal made to ASA for color densitometry. 

because the phototube response is relatively very feeble at wave- 
lengths greater than 600 millimicrons. Although somewhat broader 
in its transmission band than the rest, the red-filter combination is 
sufficiently narrow for most practical sensitometric purposes. 

The spectral characteristics of the combined system, when the 
filter control is put in the "3" and "6" positions are such as to meet the 




requirements for measuring American Standard Density Type P-2b 
It is a simple matter to remove the cover plate of the measuring 
arm and substitute other gelatin filters for those originally provided, 
should this be desired. 

400 20 40 60 80300 20 43 6O806OO204O6O80700 


Fig. 12 Log (relative response) versus wavelength for 
the "visual" filter used in the Ansco Color densitometer, 
shown in comparison with the response curve of the eye. 

The lamp voltage is controlled by a variable solenoid inductance 
and provides the required latitude of flux density. The arrangement 
permits smooth lamp operation of practically unlimited life. 


Fig. 13 is a photograph of the complete instrument. In use, the 
operator turns the instrument on with the rotary switch located at 




the left of the measuring arm and after allowing a few minutes for the 
initial warm-up makes the zero adjustment with the filter selector 
and compensating circuit controls in their appropriate positions. 
The sensitivity control, located at the right of the measuring arm, is 
then adjusted so that a calibrated reference-density specimen gives 
a reading in agreement with its assigned value. The instrument may 
then be used for the corresponding position of the "Color Black-and- 
White" switch, it being only necessary to check the zero and sensi- 
tivity-control adjustment occasionally during warm-up. It is recom- 
mended that in cases where the instrument is used daily, it be left in 
continuous operation. This minimizes the frequency at which zero 
and sensitivity checks are necessary. Since all of the components 

Fig. 13 External view of the Ansco Color densitometer. 

are operated at well below their nominal capacity, they have a long 
useful life. 

For reading color densities, with the "Color Black-and- White" 
selector switch in the "Color" position, the operator simply adjusts the 
sensitivity to give a meter reading which is in agreement with the 
preassigned reference density value using the green filter, then pro- 
ceeds to take routine readings. In the case of black-and-white 
measurements a similar check reading is made with the selector set 
switch in the "black-and-white" position and with the filter selector 
at "3" densities between 0.0 and 3.0 are then read directly. Densities 
between 3,0 and 6.0 may be read by inserting sufficient sample den- 




Fig. 14 Instrument chassis showing high-voltage 
transformer, control and rectifier tubes, lamp-voltage sta- 
bilizer, and zero-adjustment solenoid. 

Fig. 15 Liquid-density attachment in use. 

58 MONROE H. SWEET January 

sity in the beam to bring the meter reading to 3.0, while the instrument 
is adjusted to read correctly from 0.0 to 3.0, then setting the filter 
control to 6 and readjusting the zero control to bring the pointer to 
0.0 setting. 

The instrument chassis can be seen in Fig. 14. No specially se- 
lected, calibrated, or aged tubes are used. The response character- 
istics of virtually any photomultiplier tube can be compensated to give 
results which are in agreement with density values marked on the 
uniformly calibrated meter scale, by simple adjustment of the com- 
pensating circuit. The inherent stability of operation is assured by 
the basic feedback circuit. 









Fig. 16 Optical system of reflection attachment for color 

Of course the measuring arm can be used as an exploring element 
by removing the pivot shaft from the base of the arm. The inherent 
sensitivity of the instrument is between 1.0 and 0.1 microlumen at a 
density reading of 3.0. 

Densities up to 4.0 have been measured through a circular aperture 
only 0.001 inch in diameter by replacing the standard aperture plate 
with one having a smaller size opening. An attachment has been 
designed for measuring the density of liquids and is shown in Fig. 15. 

Another attachment permits the reflection densities of solid sub- 
stances to be measured in color and in black and white. The optical 
system is shown in Fig. 16. Stray light is eliminated to such an extent 
that for a typical unit less than one part in 1000 of the specularly 
reflected component of the incident beam affects the phototube when 
a first surface mirror is placed in the specimen position. 




Fig. 17 shows the head itself and Fig. 18 shows the head in actual 
use. Fig. 19 shows the spectral-energy-response product curves for 
the three-color filters used in the reflection head. 

Fig. 17 Reflection attachment for color densitometer. 

Fig. 18 Reflection attachment for color densitometer in use. 

One of the most attractive applications of the instrument is for 
automatic recording. Since the densitometer is unique in having an 




electrical output which is uniform in density it can be attached 
directly to any standard ink recorder such as the Brown high-speed 
automatic potentiometer and used as a linear automatic recording 

Fig. 19 Log (reciprocal relative response product) curves for niters 
used in reflection attachment. 

Wherever large numbers of sensitometric or similar strips are to be 
read the use of the automatic recording combination makes analysis 
faster, more convenient, more fully objective, and therefore more 
nearly free from error. 

Automatic recording systems require "smooth" modulation strips 
for satisfactory operation. In the case of the Eastman Type Il-b 
sensitometer these may be obtained by masking the steps of the ex- 


posing drum with a smooth cover plate* or by using the sensitometer 
as is and inserting a "repeating" step wedge, f 

Fig. 20 shows the instrument with a simple film-drive unit attached 
to a linear high-speed recorder. Fig. 21 shows a series of three traces 
for a color-film sensitometric strip. Each trace is recorded in the 
color corresponding to the color of light at which the measurement was 
made. Further application of the automatic recording system in the 
motion picture field will be reported at a later date. 

The speed of response is entirely dependent on the characteristics 
of the indicator since the response time of the photomultiplier tube 

Fig. 20 Color densi to meter used in conjunction with Brown high-speed 
ink recorder for automatically recording color densities. Color of the ink 
used in making the individual traces corresponds to the color of the isolation 

and associated control-tube circuits is of the order of magnitude of 
milliseconds. Therefore, in automatic recording, the speed of response 
is limited solely by that of the recorder. 


The author is indebted to several members of the Ansco Physics 
Research Laboratory : Dr. Hoerlin, Mr. Blakeslee and Mr. Alanckos, 
for their help in developing and testing the instrument, and particu- 

* United States Patent, 2,406,702, H. W. Moreall, Jr., 8/27/46. 
t United States Patent, 2,457,746, M. H. Sweet, 12/28/48. 



larly to Mr. Karl Greif for his assistance and many excellent sug- 
gestions which contributed materially toward making it a better 

Los C& 

Fig. 21 A direct trace of an actual recording of a color-film strip 
made using the equipment shown in Fig. 20 (in the original record each 
curve is drawn with an ink whose color corresponds to that of the iso- 
lation filters). The solid line was traced from the blue-ink curve, the 
broken line from green-ink, and the dotted line from red-ink. Total time 
for recording three complete traces was slightly less than two minutes 
(including loading and unloading sensitometric strip and graph sheet). 
The "pips" in the three curves correspond to a fiducial line exposed on 
the sensitometric strip itself in order to indicate the alignment of the 
three traces. 


(1) R. M. Evans, "A color densitometer for subtractive color processes,' 
/. Soc. Mot. Pict. Eng., vol. 31, pp. 194-201; August, 1938. 

(2) Monroe H. Sweet, "The densitometry of modern reversible color film," 
/. Soc. Mot. Pict. Eng., vol. 44; pp. 419-435; June, 1945. 

(3) "Report on the application of electron multipliers to spectroscopy," 
London, May 1948. J. Sci. Instr., vol. 26, pp. 53; February, 1949. 

(4) ASA Z38.2.5 (1946). "American standard for diffuse transmission 
density," American Standards Association, 70 E. 45th Street, New York, New 

(5) Monroe H. Sweet, "A precision direct-reading densitometer," J. Soc. Mot. 
Pict. Eng., vol. 38, pp. 148-172; February, 1942. 

Color Measurement of 

Motion Picture Screen Illumination 



Summary In comparing the color quality of motion picture projector 
light sources, it is of little value to judge bare screen colors, with no film 
in the gate. A complete specification of the spectral distribution of the 
radiant energy is required. Short-cut methods of evaluating and speci- 
fying this important property are described, involving the use of red, green, 
and blue filters to determine the ICI trichromatic coefficients, and, from 
these, the color temperature. A method of combining direct color meas- 
urements on the carbon arc crater from various angles of view to yield the 
screen color in any optical system of interest is also described. 

THE COLORS APPEARING on the motion picture screen when color 
film is in the projector gate are the combined result of the spectral 
distribution of the radiant energy from the light source and the spec- 
tral transmittance of the film. Useful color comparisons of light 
sources in this service thus require a much more accurate knowledge of 
their spectral qualities than can be derived from the visual examina- 
tion of bare screens, with no film in the gate. For example, the white 
color of a bare screen illuminated by an equal-energy source, with 
identical radiant intensity at all wavelengths, can be exactly matched, 
visually, by a hypothetical light source consisting of nothing more than 
approximately equal parts of sodium vapor yellow (5890 A) and a 
monochromatic blue (4860 A). Although there would be no visual 
way of telling from the identically-matched bare screens, it is obvious 
that color film would look very much better with the equal-energy 

The real test of a light source in color film projection is thus a 
visual evaluation of the picture colors finally produced on the screen. 
Independent evaluation of the light source itself is possible only in 
terms of the complete specification of the spectral distribution of the 
radiant energy as related to the transmittance of the film. Such a 
specification is very useful in many colorimetric calculations, such as 
those directed toward a determination of what a so-called standard 
observer 1 would see in any given case. In a previous paper, 2 data 

PRESENTED: October 11, 1949, at the SMPE Convention in Hollywood. 



of this sort were presented for a number of typical projector combina- 
tions, and it was pointed out that a close correspondence exists be- 
tween the spectral distribution of carbon-arc screen light and that of a 
black-body, heated to the closest color match. As applied to the 
specification of carbon-arc screen light, therefore, the so-called color 
temperature of that light has useful meaning in defining in simple 
terms a spectral distribution of sufficient accuracy for most colori- 
metric work. 

The determination of screen-light color temperature through the 
measurement of the spectral distribution of this light is, however, a 
tedious process, so that a simpler means of color temperature 
determination is desirable. Here we have found the photocell-filter 
combination suggested by R. S. Hunter 3 of value, in which compara- 
tive photocell readings are taken with specially selected red, green, and 
blue filters. These readings specify the ICI (International Commis- 
sion on Illumination) trichromatic coefficients, x, y, and z, 1 of the screen 
color in question, and these may be compared with the corresponding 
coefficients of black-bodies at various temperatures to find the one most 
nearly a match. We have confirmed Hunter's finding that it is nec- 
essary to calibrate the photocell-filter system directly against light 
sources of known color quality, similar to those to be measured. For 
instance, when our cell and filters are calibrated against an incan- 
descent tungsten standard of 2848 K (degrees Kelvin) color tempera- 
ture, we find that the red filter readings require the addition of *an 
8 per cent correction to give good values in the 5000-6000 K range of 
carbon-arc sources. 

Figure 1 shows the apparatus employed in these three-color screen 
light measurements. A photocell with a plain glass window is 
mounted on a frame, with provision to slide any one of three color 
filters in front of it. The chain drive shown permits convenient 
remote operation of the filter slide when the assembly is mounted on 
top of a pole, in the center of a full-size motion picture screen. From 
the relative readings obtained with these three filters, the trichromatic 
coefficients, x, y and z, of the screen-light color are obtained. These 
in turn are referred to a chart similar to that shown as Fig. 2, which 
is an expanded section of the ICI color diagram including a portion 
of the black-body locus, and with iso-temperature lines and uniform 
chromaticity ellipses calculated after the method proposed by Dr. 
B. D. Judd of the Bureau of Standards. 4 

As an example, the screen color with the new Hitex 5 13.6 mm carbon 




at 170 amp is. plotted at the point A. Following the slope of adjacent 
iso-temperature lines to the black-body locus gives a color temperature 
of 6250 K, directly indicated by the scale drawn along the locus. 
The divisions of this scale are drawn at the halfway points between 
the indicated temperatures (e.g., at 6225 and 6275, on each side of 
6250 K), extrapolations from color points to any point between two 
divisions being assigned the same color temperature. Our expe- 
rience in duplicating optical setups and the associated measurements 
indicates that no greater accuracy, applicable to all commercial 

Fig. 1. Photocell-Color Filter 
Assembly; used to determine 
the ICI trichromatic coefficients 
of screen light. 

Fig. 2. Section of the ICI 
Chromaticity Diagram; showing 
the black-body locus in the range 
of carbon-arc screen colors, to- 
gether with the iso-temperature 

3.0 3 

lines and the equal-color-difference ellipses used to evaluate the de- 
parture of a given screen color point from the black-body locus. 

systems of the type described, is justified, although the mathematics 
with a given set of data do permit a much closer specification. 

The chart of Fig. 2 may also be employed to give a measure of the 
difference between a screen color of interest, i.e., point A, and the 
nearest matching black-body color. According to Judd, 4 the ellipses 
of this chart define the locus of color points 10 units of least percepti- 
ble difference (LPD) removed from the color point in the center of the 
ellipse. Therefore, since the distance between point A and the 
black-body locus at 6250 K is 9/10 the nearest parallel ellipse radius, 
the Hitex carbon screen color is said to be 9 LPD from this black- 


body color. Such information is useful in evaluating the validity of 
the color temperature nomenclature in any given case. 

It is also of interest to know something of the chromatic nature of 
such a color difference. To determine this, the line joining the screen 
color point with the nearest black-body color is extrapolated to the 
spectrum locus, as in Fig. 3, to give an intercept B defining the 
spectral nature of the color difference. It is thus determined that the 
Hitex carbon screen color could be exactly matched by adding about 
2 per cent of monochromatic green radiation of 5500 A wavelength 
to the color of a black-body at 6250 K. 

Figure 3 also shows a number of other screen color points with 
various carbons and optical systems, showing in all cases a close group- 
ing about the black-body locus. Of greater significance in evaluating 
the validity of the black-body designation, however, are the curves of 
Fig. 4. These show the complete spectral distribution of radiant 
energy for the Hitex carbon screen color, together with that of the 
6250 K black-body, which most nearly matches it. The difference 
between the two curves, ordinate by ordinate between 4000 and 7000 
A, averages only 3 per cent. For most visual studies, this order of 
accuracy would seem quite adequate. 

Various proposals have been made 6 ' 7 to simplify the foregoing 
procedure by measuring only two color components of the screen light, 
associating the color temperature with the ratio of these two readings. 
Depending upon the accuracy of the calibration, this can be made a 
quite satisfactory procedure for light sources similar to black-bodies 
in relative spectral distribution, but suffers from the limitation that 
the answer is always some exact color temperature, with no indica- 
tion of the magnitude or chromatic nature of the departure from the 
black-body locus. 

Further, it should be borne in mind that the significance of these 
findings is confined to the visual spectrum, 4000 to 7000 A. From 
other studies, we know that the infrared radiation of carbon arcs is 
much less than that of the visually-matched black-body. Also, the 
ultraviolet radiation is in general much higher for arcs than for vis- 
ually equivalent black-bodies, although this is modified in marked 
degree by the absorption of the glass parts of the optics. Thus, while 
color temperature designations are useful in studies involving motion 
picture projection where the human eye is the receiving instrument, 
they must be used with considerably greater caution in other applica- 
tions, such as photography, for instance, where the recording sensi- 





Fig. 3. IOI Chromaticity Diagram; showing typical screen color points 
with relation to the black-body locus and the locus of spectrum colors. 

Fig. 4. Spectral Distribution 
of Radiant Energy; comparing 
the screen color with the new 
13.6-mm Hitex super-high-in- 
tensity carbon at 170 amp with 
that of a black-body at 6250 K. o 


tivities vary with wavelength in a manner widely different from that 
of the eye. 

So far, this paper has been concerned with the evaluation of the 
final screen color, without regard to its origin. It is also of interest to 
study the characteristics of the source and of the optics which deter- 
mine this end result. In this connection, it has been pointed out by 
Jones, 8 among others, that motion picture screen illumination is, in 
fact, an overlay of many crater images of varying magnification, 
elliptical foreshortening and orientation, as these things are deter- 
mined by the different angles with which the source is viewed by the 
elemental areas of the optical system. In this way, color and bright- 
ness variations existing at the source are smoothed over the screen 
to give a much higher order of brightness and color uniformity than 
exists over any one direct view of the source itself. This highly im- 
portant averaging effect can be demonstrated by masking the light 
collecting element in such a way that the screen light at any instant 
is confined to that delivered by a single elemental area of the collector 
element. If, then, this mask is moved about to explore the surface 
of the collector, the corresponding array of screen light distributions 
will give a sequence, one at a time, of the individual images in the 
complex overlay which constitutes the screen illumination. However, 
in a motion picture projection lamp, it is quite difficult to move such a 
masking device conveniently about inside the lamp housing while the 
projector is in operation. Fortunately, a completely equivalent 
effect can be secured much more simply, entirely outside the pro- 
jector housing, by locating a pinhole in the light beam at a suitable 
location in front of the projection lens. Just as a lens of 5 in. focal 
length will image an aperture plane 5.01 in. distant at a screen 100 
ft away, so will this same lens image the elliptical mirror in a Suprex 
arc lamp, for instance, in a plane a little less than 6 in. in front of the 
focal point of the lens. Such a mirror image is shown by Fig. 5. A 
pinhole placed in this image plane will thus limit the screen light to 
that originating from the small mirror element imaged in the pinhole 
just as effectively as if all but this area of the mirror itself had been 
blackened. In order to utilize this effect, the apparatus illustrated 
by Fig. 6 was constructed. This is adapted to locate a pinhole any- 
where over the mirror image of Fig. 5. 

Figure 7 shows a typical view of the screen light so obtained, from 
a mirror segment looking at the crater from a 65 angle. The outline 
of the aperture image has been emphasized on the negative from which 




Fig. 5. Image of the Light-Collect- 
ing Mirror of a Typical Suprex Carbon 
Arc Lamp; formed by the motion 
picture projection lens in a plane close 
to the front surface of this lens. 

Fig. 6. Pin-Hole Locating Device; 
mounted in the mirror image plane of 
Fig. 5, and adapted to restrict the 
screen light to that originating from 
mirror segment imaged in the pin-hole. 

Fig. 7. Screen Light Contribution of an Elemental Mirror Area; viewing the 
carbon arc crater from an angle of 65 , and showing the crater imaged elliptically 
in the center, with the flame light ahead of it to the right; the projection lens fails 
to pass the shell light behind the crater, which would otherwise have been imaged 
in the dark section at the left. 


this print was made. The elliptically bounded section in the center 
of the screen is the image of the high intensity crater, which, from this 
angle of view, does not completely fill the aperture. To the right, 
the less brilliant light from the arc flame in front of the crater is 
imaged. The completely dark section to the left indicates that por- 
tion of the aperture which receives no usable illumination from the 
mirror segment under study. In this particular case, light originating 
from the positive carbon shell on the side of the carbon nearest this 
mirror segment is reflected through the far side of the film aperture 
at an angle outside the cone accepted by the projection lens. 

The result of summing the illumination of Fig. 7 with that from all 
the other mirror segments is a rosette-like overlay, with crater images 
at all possible angles, and with the dim and the dark sections distri- 
buted uniformly over 360 around the center of the screen. The very 
important function of the optical system in thus averaging the non- 
uniformities of the individual mirror contributions, like Fig. 7, to give 
the uniformly white screen which characterizes a well-aligned pro- 
jection system is thus apparent. At the same time, the nonuni- 
formities in color and intensity which result from careless optical 
alignment, putting the individual images like Fig. 7 off center on the 
gate to bring in variable amounts of shell and crater light, can be 
readily visualized. 

With the condenser light-collecting optics which are commonly 
employed with larger carbons at higher currents, the same funda- 
mental considerations apply. The incomplete aperture coverage of 
Fig. 7 is avoided, however, by the use of a smaller light-collecting 
angle with a corresponding reduction in the ellipticity of the crater 
images on the aperture. This advantage is counteracted to some 
extent by the spherical aberration of the condensers, which images the 
angular views of the crater somewhat off center on the aperture. 

This breaking down of the screen light into its optical components 
suggests the reverse procedure, previously described by Jones, 8 
of measuring the light distribution over the carbon-arc crater region 
from various angles of view, and then combining these, with proper 
weighting factors, to predict the projector aperture and screen illu- 
mination with any optical system of interest. Since Jones was inter- 
ested only in evaluating light intensity, he used a conventional Viscor- 
filtered photocell in all his crater measurements. In the work de- 
scribed here, this procedure was extended to include the use of the 
clear-window photocell and the red, green, and blue filters used to 


determine the ICI trichromatic coefficients of screen light. Thus 
instead of a single trace of brightness variation across a given crater 
image, three such traces are obtained, with each one of the three 
color filters in turn. Each one of these may then be treated as Jones 
did to give a summation of the corresponding color intensity over the 
aperture with any light-collecting system of interest. 

Figure 8 shows a view of the crater image board, with a clear-win- 
dow photocell in use. The color-filter slide shown in Fig. 1 is located 
in the beam near the crater, so that any one of the three filters may 
be put into the beam. The procedure at the image board is exactly 
the same as was previously described, 8 except that now three separate 
sets of intensity variation data are obtained, one with the red, one 
with the green, and one with the blue filter in the light beam. 

Figure 9 shows such data for two angles of view of a typical carbon- 
arc crater. Through the application of the appropriate magnification 
ratio of the optical system for each angle, these color intensity varia- 
tions may be transferred to the plane of the motion picture projector 
aperture and from there to the screen. Here the appropriate weight- 
ing and summation of these with similar data from other angles of 
view give a prediction of the screen-light color resulting from the total 
overlay of crater images on the screen. 

In accumulating data with this system, a phenomenon was observed 
which is of some theoretical interest, and perhaps may have practical 
value in certain types of optical calculations as well. In correlating 
the color intensity distributions across the crater from several angles 
of view, an attempt was made to interpret these in terms of a single 
plane of origin which would produce cosine-law variations matching 
the observed behavior at various angles. No single plane could be 
found to satisfy this condition for all three colors, but it was found 
possible to secure somewhat better correlation with any one color. 
For instance, referring to Fig. 10, it was observed with the green 
radiation across the horizontal axis of the crater, that the point of 
maximum brightness appeared to originate from a point some finite 
distance inside the crater, 1.3 mm in the case described. Similar 
consideration of the blue and red components gave apparent origins 
0.5 mm and 1.7 mm inside the crater, respectively. Measurements 
of this same phenomenon, made on several other types of carbons, 
gave different absolute values, but always in the same relative order. 
This is in accord with the concept that the coolest region in the crater 
is next the crater floor, where the vaporization temperature of carbon, 

Fig. 8. Crater Image Board; adapted for tracing the red, 
green, and blue intensity distributions across the crater. 


8 4 4 SMM 

Fig. 9. Typical Crater Color Traces; 
viewed directly and from a. 40 angle. 


Fig. 10. Crater Color Intensity Dis- 
tribution versus Angle of View; show- 
ing the apparent origin of the most in- 
tense green radiation at a point 1.3 mm 
inside the crater; similarly determined 
positions for the blue and the red radia- 
tions are also indicated. 


ca. 3900 K, limits the maximum temperature to that value. As the 
distance from the crater floor increases, within the confines of the 
crater, so does the temperature of the crater gases. Such a hypothesis 
explains the occurrence of the maximum intensity for the lowest 
energy red radiation nearest the crater floor, and that of the higjhest 
energy blue radiation farther out in the crater cavity. In terms of 
screen illumination, these differences are of no practical consequence. 
They do, however, enhance our theoretical understanding of what 
goes on in the crater region, and, from that standpoint, contribute to 
the continued improvement of the high intensity carbon arc as a 
motion picture projector light source. 

In conclusion, it should be re-emphasized that any useful specifica- 
tion of the color of motion picture screen illumination must give a 
measure of the spectral distribution of the radiant energy involved. 
Trichromatic coefficients and color temperature values are of interest 
only in so far as they give such information. With carbon arcs, close 
correspondence to the black-body type of continuous radiation, with 
substantial amounts of radiant energy at all wavelengths in the visual 
region, permits the effective use of this abbreviated nomenclature. 
It is this same uniformity of spectral distribution which assures the 
effective reproduction of natural color on the motion picture screen. 


(1) A. C. Hardy, Handbook of Colorimetry, Massachusetts Institute of Tech- 
nology Press, Cambridge, Mass., 1936. 

(2) M. R. Null, W. W. Lozier, and D. B. Joy, "The color of light on the pro- 
jection screen," Jour. SMPE, vol. 38, p. 219; March, 1942. 

(3) R. S. Hunter, "Development of niters for tri-stimulus and luminosity 
measurements with barrier-layer photo-cells," J. Opt. Soc. Amer., vol. 28, p. 51; 
February, 1938; ibid., p. 179; May, 1938. 

(4) D. B. Judd, "Estimation of chromaticity differences and nearest color 
temperature on the standard 1931 ICI colorimetric coordinate system," /. Res. 
Nat. Bur. Stand., vol. 17, p. 771; November, 1936. 

(5) R. M. Bushong and W. W. Lozier, "New 13.6-mm 'Hitex' super high in- 
tensity carbon." Presented at SMPE 66th Semiannual Convention, October 11, 
1949; to be published in Jour. SMPE. 

(6) H. L. Woodward (assignee to Weston Electrical Instrument Corp.): 
"Photo-electric cell color temperature measuring device," U.S. Patent 2,462,823, 
February 22, 1949. 

(7) J. Nicholson (assignee to Photo Research Corp.): "Device for determining 
color temperatures of light sources," U.S. Patent 2,475,108, July 5, 1949. 

(8) M. T. Jones, "Motion picture screen light as a function of carbon-arc- 
crater brightness distribution," Jour. SMPE, vol. 49, p. 218; September, 1947. 

Ginecolor Three-Color Process 


Summary The basic chemical reactions, spectral characteristics of the 
dyes and types of machines utilized in the film processing are discussed in 
detail. The entire Cinecolor three-color process is described from the 
printing of negatives to the final inspection of the finished print. 

THE CINECOLOR THREE-COLOR PROCESS is a subtractive process 
whose application is found primarily in the theatrical and com- 
mercial fields where many copies from an original are required. 

The three-color process is designed for, and depends upon, three- 
strip separation negatives for its printing medium. These negatives 
may be obtained from alternate or skip-frame techniques such as are 
employed in cartoon photography and three-strip beam-splitting 
cameras; or separations made from monopack films, such as Koda- 
chrome or Anscocolor. Aside from the above, it is not the purpose 
of this paper to delve into the technique of producing negatives but 
rather to describe the print process. 

Because of the years of experience which'the company has had in 
the two-color field, the controls which have been developed, and econ- 
omies of operation which have been effected, the three-color process 
was intentionally developed along the lines of the two-color system. 
In other words, the attempt was made to develop the three-color 
method, as much as was feasible, as an extension of the two-color 

The positive raw stock utilized is the conventional and well-known 
duplitized film consisting of the usual base with color-blind positive 
emulsions impregnated with a water soluble yellow dye coated upon 
both sides of the base. It is of interest to note that this film has 
exceptionally good projection life, outlasting prints on single coated 


Assuming that three-strip separation negatives are available, the 
blue record will be referred to as the yellow printer, the green record 
as the magenta printer, and the red record as the cyan printer. The 

PRESENTED: October 11, 1949, at the SMPE Convention in Hollywood. 


first step in the process is to print two of the records simultaneously 
onto opposite sides of the positive film. While it is not essential, the 
common practice is to use the cyan and magenta printers in this first 
operation. In the same operation, the sound track is printed to the 
side of the film which is subsequently to contain the cyan image. As 
a result, the positive raw stock reaches the process department having 
latent images of the magenta component on one side of the film and 
the cyan picture component and the sound track on the opposite side. 
Because of the necessity of maintaining these two picture images in 
perfect superposition or registration, step printers rather than con- 
tinuous printers are utilized. 

These printers, as illustrated in Fig. 1, have two lamp houses con- 
nected to the film gate by means of light tunnels. As can be seen 
from the illustration, the two separation negatives are brought down 
through the film gate with their image-containing gelatin coatings 
facing each other, with the positive film sandwiched in between. 
Before reaching the film gate one of the negatives, which has been 
previously edge-notched, passes through a conventional breaker box 
whose purpose it is to actuate a light-changing device at the instant 
the change of scene occurs in the printer aperture. 

The Cinecolor printing machines utilize push-down pins located 
just above the aperture rather than the conventional pull-down pins 
which are usually present below the aperture. Because of this, the 
wear and tear on the perforations utilized for registration is minimized 
because, no matter how badly shrunk the negatives might be, the 
registration pins of the printers are caused to enter sprocket holes in 
the film which are no more than a small fraction of an inch away from 
the holes utilized for advancing the film. In this manner no punch- 
ing can occur with the registration pins and, as a result, it is the rule, 
rather than the exception, that the steadiness and excellent image 
superposition of the five hundredth copy is identical with that of the 
first copy. 

The light-change device employs a continuous loop of opaque leader 
stock which has punched in it holes of variable but predetermined size. 
This leader stock is advanced automatically by means of a solenoid- 
actuated sprocket which, in turn, is controlled by an electrical contact 
which occurs when the notch in one of the negatives referred to above 
reaches the breaker box. In other words, this might be referred to as a 
variable area type of light changer. As shown in Fig. 1, the film, 
after leaving the picture aperture, continues on down to where it is 


Fig. 1. Release printer. 



met by the sound track negative and both of these films are passed 
over a continuously rotating conventional sound sprocket containing 
the typical sound aperture which is illuminated from a lamp house 

Fig. 2. Color process machine. 


The present processing machines consist of three horizontal shallow 
troughs, one above the other, and wide enough to accommodate ten 
strands of film, as illustrated in Fig. 2. The immersion time in each 
solution and wash is held constant and the duration is controlled by 
the spacings between partitions or dams. The flow rates of the solu- 


tions and washes, as well as temperatures, are held constant. The 
wash water utilized in the process is brought up from several deep 
wells, properly filtered and passed into the main distribution tank. 
It is extremely fortunate that the temperature of this well water re- 
mains quite constant throughout the year, varying between 64.5 F 
and 65.0 F. 

Each of the ten strands of film on the processing machine, while 
operating at constant speed, is, nevertheless, independent of the 
others. In other words, each strand can be started or stopped in- 
dependently at will. The linear speed of each strand is 12 ft per 
min, making a total output capacity for each machine of 120 ft per 

As can be seen in Fig. 2, the rolls of printed film are loaded on a rack 
and the rotation of the rolls is facilitated by means of ball-bearing 
spindles which are slipped through the film roll hub and which, in 
turn, fit into a sloping slot in the rack. It should also be noted that 
the take-up reels for each strand are directly above the corresponding 
roll of film in the rack so that the entire operation of the machine can 
be handled from the loading end. The take-up reels are made of 
Bakelite and the variable speed take-up is accomplished very simply 
by having these reels rest upon two rapidly rotating Bakelite spools 
placed on adjacent and parallel-driven shafts. 

The driving sprocket for each strand is just ahead of the take-up 
mechanism and on the same level. As a matter of fact, for each 
strand there are two driving sprockets on adjacent parallel-driven 
shafts and by means of idler rollers on a swivel bracket the film can 
be held down against either of the driving sprockets. One of these 
sprockets contains 35-mm teeth and the other sprocket has 16-mm 
teeth but spaced laterally the same as on the 35-mm sprocket. This 
makes it possible to operate any strand with 35-mm, 16-mm, or 
8-mm film interspliced in any manner. This is due to the fact that, 
in the substandard field, Cinecolor utilizes 35-mm width film with 
multiple rows of either 16-mm or 8-mm perforations so that it be- 
comes necessary only to flip the pivoted bracket when a splice be- 
tween 35-mm film and either of the two substandard films appears 
at the driving mechanism. While it is perfectly possible to operate 
the machine with this single driving mechanism, it is the practice, 
however, to safeguard the machine operation with several friction 
booster drives at the two ends of the machine in order to prevent the 
possibility of strands snapping due to build up of tension. 


As can be seen from the photograph in Fig. 2, the film enters the 
machine in the top trough and progresses down the entire length to 
the other end of the machine where it passes over the end and down 
into the middle trough, where it is then going in the opposite direc- 
tion. When the film reaches the driving end of the machine, it passes 
over the end of the trough and down into the third or bottom layer, 
where it is then progressing in its original direction. When it reaches 
the far end of the machine again, it comes out of the bottom trough 
past a double set of air squeegees and then progresses up into the 
dry box which extends the entire length of the machine. The film 
then moves back toward the head end where it passes through the 
driving mechanism and onto the take-up reel. 

This very brief description of the Cinecolor process machine applies 
to all of the machines utilized by Cinecolor in its positive processes, 
with one minor exception. The three-color process involves two 
stages. Because of the simplicity of the first stage, which requires 
only a few solutions, the length of the machine permits it to be 
operated at twice the speed of the other machines, and thus to supply 
two machines utilized for the second stage of the process. To be 
more explicit, each machine is in itself a complete unit with respect 
to the two-color process, but in the three-color process three machines 
will do the work or create the output equivalent to two machines in 
the former process. 


When the film first enters the machine it is immersed in a conven- 
tional developer where the sound track, cyan, and magenta images 
are developed into silver. After a thorough wash the film is then 
passed by an air squeegee which blows off the excess moisture from the 
magenta side of the film. From here it is laid cyan down onto a solu- 
tion whose purpose is to convert the sound track and cyan images 
into a cyan pigment. While the chemical reactions involved in this 
step of the toning operations are manifold and somewhat complicated, 
they can be illustrated in rather simple terms by the following equa- 
tions : 

4Ag + 4 K 3 [Fe(CN) 6 ] - Ag 4 [Fe(CN) 6 ] + 3 K 4 [Fe(CN) 6 ] (1) 
2 Fe 2 (S0 4 ) 3 + 3 K4[Fe(CN) 6 ] - Fe 4 [Fe(CN) 6 ] 3 + 6 K 2 S0 4 (2) 


Equations (1) and (2) may be combined to show the complete re- 
action as follows : 

4Ag + 2 Fe 2 (S0 4 ) 3 + 4 K 3 [Fe(CN) ] -> 

Ag 4 [Fe(CN) 6 ] + Fe 4 [Fe(CN) 6 ] 3 + 6 K 2 SO 4 (3) 

The above equations indicate that the reaction takes place prima- 
rily between the silver of the image, potassium ferricyanide, and a 
ferric salt, and that the end products of the reaction consist of silver 
ferrocyanide, ferric ferrocyanide (Prussian blue) and potassium sul- 

In addition to these reactive agents the toning solution of course 
contains other materials which, because of ionization equilibria and 
the formation of complex ions, can control the availability of the 
reactive ions and consequently the quantity and character of the 
Prussian blue deposit which is formed. It is perfectly possible to 
control contrast and the degree of dispersion of the ferric ferrocyanide 
deposit by varying quantitatively and qualitatively the composition 
of the toning solution. It is possible to form a coarse grainy agglom- 
eration which produces bad grainy effects on the screen as well 
as the destruction of resolution and definition, or it is possible to 
produce a colloidally dispersed deposit which is highly transparent, 
free from grain, and having high resolution characteristics. It is 
also possible to go beyond this point and produce such a high degree 
of dispersion that bleeding takes place, causing once again the loss of 
resolution. Once all the factors are known and properly controlled, 
it is a simple matter to form a cyan image which has excellent grain 
and resolution characteristics. In addition to this, the spectral qual- 
ity of this type of image is good from the standpoint of three-color 
reproduction, as may be seen by reference to Fig. 3. A more complete 
discussion of the spectral characteristics of this image will be given 

It will also be noted from the above equations that the silver which 
forms the original cyan image has at this stage of the process been 
converted to an insoluble silver salt, namely, silver ferrocyanide, 
which salt is both light insensitive and spontaneously developable 
if brought into contact with a developing solution. 

When the cyan image has been completely converted in the solu- 
tion referred to above, the film then passes into a wash where the un- 
reacted toning solution is completely removed from the film, after 
which the film is immersed in another solution whose purpose is to 




convert the silver ferrocyanide to silver bromide. At this stage the 
silver of the original cyan image is in the form, namely, silver bromide, 
where it was prior to the printing operations, with the exception that 
the original silver bromide crystalline structure has been destroyed. 
This reformed silver bromide is neither subject to spontaneous 
development nor is it particularly light sensitive. 










/ / 












































i i 




\ / 










/ \ 











v > 




s ' 











00 20 40 60 60 500 20 4 

60 80 600 20 40 60 80 700 

Fig. 3. Spectral-density characteristics of yellow, magenta, and 
cyan components. 

It is possible, however, by controlling the character of the ferric 
ferrocyanide deposit surrounding these particles of reformed silver 
bromide to produce a degree of photographic sensitivity which is 
higher than that of the original silver bromide grain which existed in 
the raw unprinted form. By controlling further the character of the 
ferric ferroeyanide deposit it is possible to vary the sensitivity of the 


reformed silver bromide in such a manner that this sensitivity is 
increased in direct proportion to the decrease in exposure brought 
about by the masking effect of the ferric ferrocyanide image. In 
other words, if one were to flash expose the film containing a cyan 
image the amount of latent image present in every part of the picture 
would be constant and this constant image could be brought up into 
silver by subsequent development just the same as though there were 
no cyan image present. This is an important consideration in view 
of the fact that later on in the process it is desirable and necessary 
to produce an additional image on the cyan side of the film without 
interference from the cyan image already present. 

Upon leaving the solution last mentioned, the film is washed again, 
properly hardened, and given a final wash before it enters the dry box 
and is subsequently removed from the machine. The carefully con- 
trolled drying operation produces no appreciable shrinkage in the 
film because of the protection afforded the base by the coatings on 
each side. When the film comes off this first machine it has on one 
side an unfixed photographic emulsion containing a silver image of 
the magenta component and on the opposite side a cyan image im- 
bedded in a complete photographic emulsion whose characteristics are 
such that the effective sensitivity of the entire surface is constant 
irrespective of the presence of that image. 


The next step in the process is the printing of the yellow component 
through the yellow printer negative. At present, this printing opera- 
tion is accomplished on machines similar to those described above. 
When this operational step is completed the film is ready for its final 
processing. It should be remembered that the yellow dye with which 
the emulsions of the film were originally impregnated was leached 
out of the film in the first developing stage so that in the second print- 
ing operation some of the light to which the film is sensitive has pene- 
trated through to the opposite side. In order to overcome this dele- 
terious effect, the film is floated on the developer, the first solution 
of the second processing stage. This brings up the yellow compo- 
nent in silver on the cya'n side of the film and the portion of the image 
which has penetrated through to the opposite side is allowed to die 
as a latent image. 

It should be noted at this point that the floating operations involved 
in this process present no problems due to the fact that it is possible 


to maintain high surface tension characteristics in the corresponding 
solutions. It is only on rare occasions that any trouble is encoun- 
tered and this is due usually to raw stock defects. After leaving the 
developer, the film is washed and then proceeds into a hypo solution 
where the undeveloped silver bromide is dissolved and removed from 
both sides of the film. Following this, the film is again washed. Next 
in the process a bleach or oxidizing solution is used to convert the 
silver in the yellow and magenta component images to a dye mordant. 
Like the cyan toning step, this one involves a group of somewhat com- 
plicated chemical reactions which can be condensed and stated quite 
simply in the following equation : 

2Ag + I 2 -> 2AgI (4) 

As can be seen from this Equation (4), the bleaching solution contains 
iodine as a principal reactant, which combines with the silver of the 
image to form an insoluble silver iodide image that has the property of 
absorbing basic dyes. As in the case of the cyan toning operation, 
by controlling the concentrations of several of the constituents in this 
bleaching bath the degree of agglomeration or dispersion of the silver 
iodide deposit can be varied at will and controlled. It is quite evident 
that if the deposit is coarse the final image will have a high degree of 
opacity. This is due to the mordant itself and to low color saturation, 
not only because of the neutral component introduced by the mordant, 
but also because of the low saturation of the image with dye due to a 
high volume-to-surface ratio of the silver iodide pa.rticles. 

On the other hand, the mordant image can be made so highly trans- 
parent that it is hardly visible prior to the dyeing operation, and, since 
the deposit consists of extremely small particles and the surface-to- 
volume ratio is high, the amount of dye absorption is very much 
greater. In this case, with the mordant having practically no opacity 
and with a high dye concentration in the image, the saturation of the 
color components is excellent. As in the cyan toning step, it is also 
possible to overshoot in this direction so that bleeding can occur, 
which of course destroys resolution. 

When the bleaching step just referred to has been completed and 
the film has been washed, the magenta side of the film is blown off by 
air squeegees and the film is then floated on the yellow dye. Upon 
emerging from this solution, the film is washed again for a short period 
of time and is then passed by air squeegees to remove the excess 
moisture from the yellow side of the film in order that the magenta 


side may be floated upon the magenta dye solution. After a final 
wash, the film is run through the dry box and emerges as a finished 
three-color print. 


The characteristics of the Cinecolor three-color process may be 
best demonstrated by reference to Fig. 3 which shows the spectral 
density characteristics of the three components balanced to equal 
analytical densities. 1 It is of interest to note that the peak density 
of the yellow component occurs at wavelength 445 m^t, which corre- 
sponds closely to wavelength 440 m/z, which is commonly weighted 
because of the spectral sensitivity of the human eye, and that the peak 
density of the magenta component corresponds to 540 m/x, which 
is also weighted for the same reason. 

In the case of the cyan component, however, the density continues 
to rise beyond the weighted wavelength of 640 m/x into the infrared. 
This characteristic is responsible for the high fidelity reproduction 
obtainable with Prussian blue sound tracks when used in conjunction 
with the caesium photocell whose peak sensitivity is in the infrared. 
Still concentrating on the cyan component, it will be noted that the 
density at 540 m/* is slightly higher than that at 440 m/z, which is the 
worst defect in the entire process. The degree of unbalance at these 
two wavelengths, however, is negligible, as evidenced by the high 
fidelity of color reproduction in this system. The result to be ex- 
pected from this small defect is a slight reduction in the brilliance 
of green objects in the picture. This, however, becomes somewhat 
advantageous with respect to the possibility of obtaining good sound 
reproduction utilizing the potassium S4 photoelectric cell in sound 
reproducers. The over-all peak sensitivity of this type of cell, when 
used with incandescent exciter lamps, occurs in the green portion of 
the spectrum and by actual tests it has been found that the Prussian 
blue sound track with a slight modification of print density or sound 
negative gamma is just as satisfactory with this as with the caesium 
type cell. 

It can be said, therefore, that the cyan component of the Cinecolor 
three-color process, while not perfect, is satisfactory. In the case 
of the yellow and magenta components, it can be observed in Fig. 
3 that these two are excellent. The high degree of spectral quality of 
these three components may be illustrated in a different manner, 
namely, by observing the fact that the integral densities 2 of the three 


components at their corresponding weighted wavelengths are in al- 
most perfect balance when the analytical densities of the three com- 
ponents are identical. In other words, as shown in Fig. 3, the analyt- 
ical densities of the three components are adjusted to the value 1.45. 
At the same time the integral density of the yellow component at 440 
m/i is 1.24, that of the magenta component at wavelength 540 m/x 
is 1.24, and that of the cyan component at wavelength 640 mju is 

It may be of interest to state at this point that the conversion 
factors from densities, as read on the E.R.P.I. densitometers, to 
analytical densities are 1.20, 1.30, and 1.40 for the cyan, magenta, 
and yellow components, respectively, when read through the Kodak 
Wratten 29, 61N, and 49 filters. As a result of the excellent balance 
between the E.R.P.I. densities of the three components and also be- 
tween the integral densities at the weighted wavelengths when the 
three components are adjusted to equal analytical densities, it is 
evident that the process is capable of reproducing very accurately 
colors as well as neutral values. For example, if it were required to 
photograph a scale of reds which would be composed of yellow and 
magenta on the color print, the dominant wavelength of the scale 
should remain constant throughout. If it were necessary to maintain, 
for example, a low yellow integral gamma in order to compensate 
for a small density peak in the blue portion of the spectrum resulting 
from an imperfect magenta component, then a scale of reds would 
appear either red in the low densities and magenta in the higher den- 
sities or red in the higher densities and orange in the low densities. 
In addition to this, if the same condition prevailed as stated in the 
previous sentence, it would be impossible to reproduce accurately a 
scale of yellow densities since the heavier yellow densities would be 
lacking in saturation or conversely the lower densities would be too 

In order to obtain a system which is capable of giving excellent 
picture reproduction both as to neutral and colored objects, it is essential 
that the spectral density characteristics of all of the components be of 
such a nature that a simultaneous balance will occur not only between 
the analytical densities or gammas but also between the integral 
densities or gammas. At the same time, the conversion factors be- 
tween analytical and integral density or gamma should be as close to 
unity as possible. It has already been noted above that the integral 
densities are in almost perfect balance for equal analytical densities 


and the ratios between analytical and integral densities are 1.45:1.24, 
1.45:1.24, and 1.45:1.23, for the yellow, magenta, and cyan compo- 
nents respectively. These ratios are approximately 1.17. Qualita- 
tively it can be seen from the curves in Fig. 3 that there is very little 
density of the cyan and magenta components in the blue portion of 
the spectrum, very little density of the yellow and cyan components 
in the green portion of the spectrum, and very little density of the 
yellow and magenta components in the red portion of the spectrum. 
This makes for brilliance and high saturation of color, when desired. 
The wavy neutral curve shown plotted above the spectral density 
curves in Fig. 3, when evaluated by means of the trichromatic 
coefficients 3 indicates that the neutrals should appear to be an excel- 
lent visual gray, and this fact is supported by actual screen tests. 
The trichromatic coefficients for a high intensity projection arc have 
been found to be: x = 0.3408 and y = 0.3583. Trichromatic coeffi- 
cients for the arc as seen through the Cinecolor three-color neutral 
are: x = 0.3533 and y = 0.3423. Taking the point located by the 
trichromatic coefficients for the arc as the white-point, the dominant 
wavelength of the visual gray is 538 C millimicrons, and the purity 
6.25 per cent a most adequate neutral. It can readily be seen that 
these values can be plotted as a point very close to the center of the 
Maxwell curve. 


(1) R. M. Evans, "A color densitometer for subtractive processes," Jour. 
SMPE, vol. 31, pp. 194-201; August, 1938. 

(2) M. H. Sweet, "The densitometry of modern reversible color film," Jour. 
SMPE, vol. 44, pp. 419-435; June, 1945. 

(3) A. C. Hardy, Handbook of Colorimetry, Massachusetts Institute of Tech- 
nology Press, Cambridge, Mass., 1936. 

New Projection Lamp and 
Carbon-Feed Mechanism 



Summary This paper describes a new carbon-arc projection lamp em- 
ploying an electronic system to provide automatic control of the carbon posi- 
tions. Each carbon-feed mechanism is separately driven by an alternating- 
current motor actuated by an electronic impulse generator. Feeding speed 
of the carbons is controlled by varying the resistances through the generator 
control circuits to adjust the number of impulses supplied per minute to the 
alternating-current motors according to the carbon-consumption rates. A 
polarized directional electromagnet controls the position and shape of the tail 

np HEATER MOTION PICTURE PROJECTION requires a highly concen- 
JL trated, high-intensity light source to give the greatly enlarged 
image sufficient screen brilliance. The carbon arc is the only light 
source so far developed capable of supplying the necessary screen illu- 
mination. In the past forty years or more this arc source has under- 
gone many changes both in the lamp mechanisms and in the carbons 
themselves, with resultant marked improvement in the quantity and 
quality of the light produced. 

In order to provide the constant screen light necessary to satisfac- 
tory projection, close control of the carbon positions is essential par- 
ticularly in the reflect or- type high-intensity carbon-arc lamp. The 
lamp must provide three main features: (1) The positive crater must 
be held in exact focus. (2) The carbons must be fed together at the 
same rate at which they are consumed. (3) A magnet must be pro- 
vided to minimize the effects of the magnetic field set up by the direct- 
current supply to the arc and to stabilize the position of the arc tail 

In the direct-current arc, the rate of carbon consumption varies 
according to the type and diameter of the carbons used, the current 
at the arc and the length of the arc gap. Since the positive carbon is 
consumed at a much faster rate than is the negative carbon, conven- 
tional lamps employ two feed mechanisms, one driving the positive 
carbon carrier and the other, the negative. Usually the two carriers 

PRESENTED: April 8, 1949, at the SMPE Convention in New York. 


88 J. K. ELDERKIN January 

are propelled by feed screws driven from a single direct current motor 
through speed reduction gears, clutches, ratchets, and cams. The 
cam or ratchet mechanism provides a means of regulating the travel 
speed of the individual carriers, while a rheostat regulates the overall 
speed of the single driving motor. 

In general, these motor-driven feeding mechanisms do not maintain 
the crater of the arc in correct focus for any considerable length of 
time, and, as the crater face of the positive carbon wanders from its 
exact focal point, the light on the screen changes in color and inten- 
sity. Further, these mechanisms do not maintain a constant arc 
length, slight changes in the arc current or voltage resulting in changes 
in the motor speed which cause the feed mechanism to run too fast 
or too slow. Manual adjustment of the speed of one feed mechanism 
is unsatisfactory since such adjustment also affects the speed of the 
other carbon-feed mechanism. 

To overcome these difficulties a new lamp has been developed which 
uses a separate drive-and-feed mechanism for each of the carbon 
carriers. Each carrier is driven by a nearly constant speed alternat- 
ing-current motor energized intermittently by pulses fed from inde- 
pendent impulse generators. Any desired feeding speed can be ob- 
tained merely by adjusting the number of impulses per minute fed to 
the motor, so that almost micrometer adjustment of speed can be 
obtained, making it possible to hold the arc gap constant and main- 
tain the arc in exact optical focus with the reflector-and-lens system 
to suit any required consumption rate of either positive or negative 

Fig. 1 shows the complete positive carbon carrier with the cover 
removed. Fig. 2 shows this carrier in alignment with the negative 
carbon carrier, also with cover removed, and illustrating the motor, 
motor reduction gear, worm screw, carrier with side clamp, and car- 
bon clamp. The carbon holders are full-floating and self-aligning, 
and are provided with fixed-pressure spring-tension clamps. The 
negative carbon guide and holder is adjustable both vertically and 
horizontally and, generally, is positioned so that its center is hi line 
with the bottom edge of the positive carbon crater. To strike the 
arc the carrier is brought forward by pushing in the handle at the end 
of the worm screw. After the arc is struck, the carrier springs back 
to provide the proper arc gap. The motor drives the worm shaft 
through a friction clutch. For manual adjustment of the carbon 
positions, the worm shaft is turned against the friction of the clutch. 




In order to secure the desired precision in carbon-feed control, it 
was necessary to develop a very accurate electronic impulse generator 
for supplying the energizing impulses to the alternating-current feed 
motor, and to incorporate a means for readily varying the number of 
impulses per minute between the upper and lower speed requirements 
for feeding the carbon being consumed. This generator is energized 

Fig. 1 

Fig. 2 

from the 110-volt illuminating-current supply. It employs a small 
Thyratron tube, Type 2D21, in circuit with capacitors and resistors 
in such manner that the time cycle during which the current flows 
and ceases to flow may be regulated at will, simply by increasing or 
decreasing the amount of resistance in the control circuit. By merely 
turning the knob of the variable resistor clockwise or counterclock- 
wise, the number of impulses from the impulse generator will be 




increased or decreased. The impulse generator delivers its impulses 
to the actuating coil of a relay which by the opening and closing of 
its contacts intermittently connects the alternating-current motor 
with the 110-volt alternating-current supply. Any arc-voltage 
change does not alter the feeding speed as is the case when a direct- 
current motor is used, because the alternating-current motor is actu- 
ated from an entirely independent circuit. Thus the actuation of 
the alternating-current motor and associated feed mechanism can be 
controlled perfectly and accurately in order to feed the carbon forward 
at an exact rate. By the use of one of these impulse generators for 


Fig. 3 

Fig. 4 

operation of the negative-carbon feeding mechanism and another for 
the positive-carbon feeding mechanism, a separate and accurate 
control of each is obtained. 

Fig. 3 is a view of one of the electronic timers showing the Thyra- 
tron tube, transformer, capacitors, and resistors in circuit, the relay 
for opening and closing the motor circuit, and the connector plug, 
which automatically makes all electrical connections. 

Fig. 4 is a schematic circuit of the electronic impulse generator and 
Fig. 5 is a schematic of the electrical circuits employed in the pro- 
jection lamp. 

Fig. 6 shows the relay which prevents the lamp feeding mechanism 
from operating when there is no arc between the carbons. This is 




accomplished by using a series coil in the arc-supply circuit to actuate 
the relay, the contacts of which open and close the motor circuits. 

In all arc lamps of this type a magnetic field is set up from the 
direct-current supply to the carbons which has a marked effect on the 
burning characteristics of the arc itself. Heretofore, it has been the 


Fig. 5 

Fig. 6 




practice to place either a permanent or electromagnet in the lamp- 
house, positioned in a manner so that this inherent magnetic field is 
partially neutralized to give a better burning characteristic to the arc. 
This type of arc is called a "Suprex" arc and also a "Simplified High- 
Intensity" arc. While it is a great improvement over the previous 

Fig. 7 

Fig. 8 

Fig. 9 


type of arc known as low intensity, it is still not the ultimate to be 
desired for very large theaters. 

In the lamphouse used with the new feeding mechanism described 
here, a polarized directional electromagnet is provided. This sup- 
plies a strong polarized magnetic effect which causes the arc to burn 
with a very long and narrow tail flame rising straight up at right 
angles to the arc and with little enveloping flame. Fig. 7 shows the 
shape of the tail flame as compared with that of the "Suprex" arc 
shown in Fig. 8. The magnet is shown in Fig. 6 (bottom center). 
The horn of the magnet has a target at its upper end and it may be 


Fig. 10 

adjusted vertically or horizontally on its axis. Once the magnet has 
been initially set up, no further adjustment is necessary, through the 
complete amperage range of the arc (40 to 90 amperes). 

Other features of the lamp design include a warning light which in- 
dicates when the carbon has burned beyond the point at which it will 
last through another reel of film. Both carbon carriers also have 
scales and pointers indicating remaining burning life. 

Fig. 9 shows the working side of the lamp with the door open and the 
flame shield raised. Fig. 10 is a view through the back door of the 


lamp showing both electronic timers with covers removed. Fig. 11 
is a side view showing control knobs for adjusting the carbon feeds. 
All the electric wiring in the lamp is shielded and the wire covered 
with asbestos braid. The reflector is elliptical, 14 inches in diameter 
for currents to 75 amperes and 16 inches for currents from 75 to 90 
amperes, with a center hole 2 l / 2 inches in diameter through which the 

Fig. 11 

negative carbon, carbon holder, and carbon guide pass horizontally. 
A flame shield is provided which protects the reflector when the arc 
is struck, and moves out of the way when the light gate is opened to 
pass the light to the projector. The positive carbon guide is pro- 
vided with a removable chute or shield which carries the copper drip- 
pings from the carbon into an easily removable receptacle. 

Industrial Sapphire in 
Motion Picture Equipment 






Summary This paper is an endeavor to call attention to an engineering 
material which has the most favorable coefficient of friction in relation to 
film emulsions and bases among known commercial products. 

Characteristics of sapphire in fine watches, the development of industrial 
sapphire, and potential motion picture equipment applications are dis- 
cussed briefly. The Auricon camera film gate with regard to the use of hard 
contact points and the mounting of sapphire contact surfaces is described and 
illustrated. Physical characteristics of sapphire are shown with charts and 
performance data. Diamond lapping compounds and the metal bonding and 
flame forming of sapphire are briefly mentioned. 

SYNTHETIC SAPPHIRE JEWELS have the most favorable coefficient 
of friction in relation to film emulsions and bases among known 
commercial products. This friction characteristic is due to sapphire's 
monocrystalline structure and complete lack of porosity, coupled with 
extreme hardness and resistance to abrasion. 

The value of these physical characteristics has been demonstrated 
for over a century in the jeweling of fine watches, first with natural 
stones, and of late years with synthetic sapphire which is comparable 
in hardness and not subject to the flaws found in the natural product. 

Every individual who possesses a fine watch carries from seven to 
twenty-one pieces of sapphire on his person daily, with scarcely a 
thought that these sapphire jewels represent the most accurate and 
certainly the most dependable and enduring components of the time- 
measuring mechanism. 

Many engineers are familiar with the amazing magnitude of bal- 
ance oscillations and escapement action of a watch, but few realize 
that, in spite of 432,000 impulse and locking cycles each day, the sap- 
phire pallet stones show no wear in a lifetime of service. 

Industrial sapphire, or corundum, has been man-made for some 
fifty years and until the last few years was largely a European product 

PRESENTED: May 20, 1948, at the SMPE Convention in Santa Monica. 



limited in scope of application by the small size of available raw ma- 
terial. Under the impetus of war demands for jewel bearings, indus- 
trial or synthetic sapphire is now produced commercially in the 
United States. 

Industrial sapphire is an artificially grown crystal, formed through 
an oxygen-hydrogen flame. The alumina, melting and impinging on 
a refractory pedestal in a high-temperature furnace, grows as a boule 
or a rod according to the rate at which the pedestal is withdrawn. 
Rods are commercially grown over a range of diameters to slightly 
over 0.250 inch and between 2 and 6 inches in length. Boules are 
roughly 0.750 inch in diameter, limited in length, and average about 
200 carats. Tubing, disks and other shapes have been produced ex- 
perimentally and boules as large as 800 carats have been grown. 

Fabrication is primarily by diamond-charged saws and laps, with a 
relatively new technique of flame forming and flame polishing finding 
increasing application. 

The question now is whether industrial sapphire can find a place 
for itself in motion picture equipment. The potential motion picture 
applications fall roughly into four classifications: (1) Where the 
material is urgently needed because of partial failure of existing 
components, resulting in danger of film damage. (2) Where film 
footage is sufficient to indicate marked economy in terms of greatly 
increased life of film or machine components. (3) Where an addi- 
tional safeguard could be afforded against film-base scratches and 
emulsion pickup. (4) Where a design advantage is afforded in the 
mechanism entirely apart from contact of sapphire itself. 

Let us consider current research applications of sapphire in the 
film gate of the 16-mm "Auricon-Pro" single-system sound-on-film 

The Auricon camera has a claw intermittent and it is necessary 
that the pressure plate maintain the negative in a steady position as 
the claw leaves the perforation and throughout the time of shutter 
opening in order to secure a steady in-register picture. It is also 
necessary that the pressure plate take out the natural film curl and 
hold the film flat and perfectly in focus at the aperture. This means 
that one must exert confining forces on the film during shutter open- 
ing, and since the pressure plate is of the spring-loaded, constant- 
pressure type, those same forces are present during the pulldown cycle. 
To prevent scratching of acetate film base and picking up of emul- 
sion the friction coefficient of the gate must be favorable to the film. 




Fig. 1 16-mm "Auricon-Pro" single-system sound-on-film camera. 

Like many other devices in the 
industry, the " Auricon-Pro" 
camera film gate has gone 
through years of evolution and 
redesign. These changes have 
been influenced by the fact that 
simplicity of design required the 
film to be confined with appreci- 
able pressures throughout the 
entire intermittent cycle, and 
placed severe requirements upon 
materials used in the film-gate 

Within the past year Berndt- 
Bach redesigned the "Auricon-Pro" camera gate to permit the use 
of hard contact points in the aperture plate so located that these 
contacts do not touch either the picture or sound track area of 
the film emulsion. In the back pressure plate a single contact 

Fig. 2 Auricon aperture 
sure plates. 

and pres- 


is located directly behind the center of the aperture. The film 
is fully confined and in focus at the actual aperture area. The 
eleven aperture plate contacts are located to secure distribution 
of the pressure forces over the entire aperture plate, in addition to 
providing film support at the aperture opening and at the entering 
and leaving positions of the pulldown claw. 

At the time of this redesign we believed the use of highly polished 
hard-chrome-steel balls offered a favorable material. Initial tests 
proved satisfactory, but it became evident that these balls in contact 
with film emulsions developed pits and corroded to an extent that 
nullified the advantages of their original mirror polish and hardness. 

The hard-chrome-steel balls were then replaced with balls of a 
special stainless steel of the same diameter, to secure improved cor- 
rosion-resistant qualities. Realizing that these stainless-steel balls 
woul'd be subject to surface abrasion from acetate film base and emul- 
sion regardless of original fine surface finish, the balls were lapped and 
polished after assembly to provide a tiny flat on each and slightly in- 
crease the contact area. This application has produced the most 
satisfactory Auricon camera film gate thus far obtained with metals, 
and patent applications have been allowed on this construction. 

Concurrently, investigation of the possible uses of industrial sap- 
phire in this application was undertaken, since the redesign was based 
on an endeavor to secure point or very small area contacts of extremely 
hard, nonporous material located in emulsion areas which would not 
encroach on picture or sound track. 

Sintered carbide materials, which offered excellent hardness and 
resistance to film abrasion, were ruled out on the two counts of 
porosity and unfavorable friction coefficient. 

Application of industrial sapphire was contemplated from the in- 
ception of the program. Short of diamond, sapphire offers the best 
combination of hardness, nonporosity and friction coefficient quali- 
ties, and the additional advantage of a monocrystalline structure. 

Clear sapphire balls of approximately 3 /32 inch in diameter were in- 
serted in the Auricon-Pro camera gate. (There is no practical dif- 
ference between the clear and the ruby-red sapphire, aside from 
color.) With the single-point ball contact of the back pressure plate 
directly behind the center of the aperture, no difficult sapphire-ball- 
mounting problem exists. However, with the eleven ball contacts of 
the aperture plate it is not only desirable that the multiple inserts 
cumulatively form a perfectly flat plane but, in the case of those im- 


mediately surrounding the aperture, a must. This can be main- 
tained by carefully pressing in the contacts to uniform height or, 
preferably from a production standpoint, by flat-finish lapping with 
fine diamond compound. In this case even a very slight lapping to 
produce a small polished flat on the spherical surface considerably 
increases the contact area and provides a safer margin for film gate 
pressure-plate loading. 

500 1000 I5OO 20OO 3000 4000 5000 6000 7OOO 

Sapphire has a hardness of 9 on 
the Moh scale and of 1500-2000 

HARDENED STEEL 400-800 on the Knoop scale. Its hard- 
ness and toughness exceed those 
GLASS 300-600 of any other man-made material 

Fig. 3. Hardness chart. 


Sapphire offers the motion picture industry some remarkable 
physical characteristics, the most favorable of which is its friction 
coefficient in relation to film. Use of the material has been limited 
to a very few applications in the industry, of which we might men- 
tion the use of sapphire jewel bearings and film-edge guides. 

Table I will serve to suggest, merely on the basis of physical char- 
acteristics, where sapphire will find worth-while application. We see 
no need to suggest that sapphire indiscriminately replace existing 
components. Where no present vital problem exists there is little 
need to seek an immediate solution. However, in contemplating 
redesign of existing equipment, or development of new devices, the 
designer can hardly ignore consideration of so favorable an engineer- 
ing material. 

A note to the designing engineer would not be amiss, in calling to 
his attention that sapphire becomes a fragile material when subjected 
to shock of various natures. Sapphire subjected to severe mechani- 
cal shock might lead one to believe that data on compressive strength 
and elastic characteristics is in the nature of a snare and a delusion. 
We emphasize this despite personal tests wherein ruby jewels of 0.015 
inch in thickness were placed on a steel plate and vigorously pounded 
with a pyralin hammer, without breakage. 



(Technical Data) 


Properties of Synthetic Sapphire 

Hardness and Toughness 

Chemical Formula A1 2 O 3 


Melting Point 2050 C-3722 F 

Hardness ness 

Crystal System Hexagonal 

(Knoop (Rela- 

Chemical Resistance Not affected by 

Scale) tive)* 

any known chemical at room tem- 

Diamond 6,000-6,500 2 

Modulus of Rupture 59,000 to 125,- 

Boron Carbide 2,200 4 
SAPPHIRE 2,000 3 

000 pounds per square inch approxi- 

Tungsten Car- 

Compressive Strength 300,000 pounds 

bide 1,000-1,500 2 
Hard Chro- 

per square inch 
Water Absorption 
Density, Ib. per cubic inch . 144 

mium Plate 850-900 1 
Stellite 500-800 1 
Steel, Rockwell 

Specific Gravity (Water = 1) 3.99 

C65-67 791 1 

Elastic Modulus Steel 30,000,000 
Sapphire 55,000, 000 


Refractive Index 1 . 757, 1 . 765 

T 11 1 A. 

Parallel to 

Thermal Conductivity 0.007 cal.sec. -1 

axis 710 

cm.-^C" 1 
Thermal Coefficient of Expansion 3 . 3 

lar to axis 790 4 

X 10-6 per degree Fahrenheit aver- 

Steel, Rockwell 


C47 496 1 

Dielectric Constant 7 . 5-10 . 
Stability on Aging No change 

Fused Quartz 475 4 
Crown Glass 420-470 5 

Specific Heat 

*Comparative order of toughness, 

highest being number one. 

Silver 0.05 

Copper 09 

Coefficient of Friction 

Steel 10 

Against Steel 

Glass 0.16 

Sapphire 0.14 


Graphite 0.16 

Porcelain.. 0.26 

Leaded Brass 0.19 

Shock of another nature to which sapphire should not be subjected 
is thermal shock, or sudden changes in temperature. This should be 
taken into consideration when designing with a view to using the new 
metal-bonding techniques. Until recently, solder recommendations 
in connection with soldering metal-bonded sapphire shapes have been 
limited to low-temperature solders. However, silver soldering of 
metal-bonded sapphire shapes is becoming practical. 


Extended research has been undertaken in the technique of flame 
forming and finishing of sapphire. In the process, sapphire is heated 
to a very high temperature in an oxidizing atmosphere and can be 
formed by ' 'slipping" the planes of the crystal. In spite of very dras- 
tic deformation in shape the sapphire still maintains a monocrystal- 
line character. During this operation the surface of the material 
flows to form a very high surface finish. Thus far, flame forming is 
limited to rod stock under 0.125 inch in diameter. Pigtail thread 
guides formed by 360-degree bends in sapphire rods offer an excel- 
lent illustration of the possibilities of flame forming. 

Applications of sapphire to motion picture equipment are few in 
number and too limited to enable one to make an accurate estimate 
of the full worth of this newly advailable material at the present time. 
Friction coefficients, resistance to abrasion, hardness, physical stabil- 
ity, and other vital qualities have been demonstrated through applica- 
tions in other industries. In sapphire we are dealing with an engi- 
neering material of known characteristics and increasing availability, 
in which the field of potential application is still largely unknown. 

One of the most interesting potentials appears when we think of 
film as an abrasive textile, abrading to a greater or less degree each 
surface with which it comes in contact during its life span, and in turn 
being itself abraded through the deterioration of those same surfaces. 
The thought of introducing sapphire surfaces at critical or strategic 
locations along thatjfilm path^then becomes interesting indeed. 


CHAIRMAN WATSON JONES: Are there any questions? 

MR. SAUTER: With reference to precision bearings for mechanical engineering 
use, what class of accuracy compared to standard precision bearings could we 
obtain in ball bearings? 

MR. CHRIS WAGNER: Balls are being made to an accuracy of twenty-five 
millionths sphericity and 0.0002 inch for diameter. The diameter tolerance 
could be reduced further if necessary for application in ball bearings. 

MR. SAUTER: In high-speed work what type lubricant would you need? 

MR. WAGNER: The practical value of sapphire in high-speed work is the 
elimination of lubricant. It will run dry in contact with steel races or with sap- 
phire races. 

MR. SAUTER: How does the cost compare with high-precision regular ball 

MR. WAGNER: The cost is higher. In Vis balls the cost is about sixteen cents 
per ball, and as the balls get larger the cost goes up. However, it does offer pos- 
sibilities in engineering work where the ball bearings are subject to extreme speed 
or to corrosion. 

Report of SMPE 
Standards Committee 


IN HIS REPORT 1 as retiring chairman of the Standards Committee, 
F. T. Bowditch offered several suggestions for improving the ef- 
fectiveness of the Committee. Briefly, he proposed that the develop- 
ment of standards could most properly be done by the several engi- 
neering committees of the Society and that the past practice of organ- 
izing subcommittees of the Standards Committee to prepare such 
standards should be discontinued wherever possible. Further, he 
proposed that the Committee on Standards should be so organized 
that it functions primarily as a policy-making group, but it should 
also be competent to consider the over-all effect of standards proposals 
in relation to the industry and to related standards. 

These recommendations were promptly adopted by the present 
chairman and, as a result, the Standards Committee has been func- 
tioning in substantially the manner envisioned by Bowditch. Since 
the new plan obviously called for close co-operation with each of the 
engineering committees and since the chairmen of such committees 
are obviously qualified in their respective fields, it was felt that a first 
step toward a representative policy-making group would be to invite 
these chairmen to serve as members of the Standards Committee. 
The second step was to include as members the chairmen of the sev- 
eral ASA sectional committees having interests closely related to the 
motion picture industry. A third and final step was to solicit the 
participation of the Motion Picture Research Council and of only a 
few of the many members of the Society who, either because of special 
knowledge or long experience with standardization problems, could 
be of outstanding assistance to the committee. The Engineering Vice- 
President approved this plan, and the present committee has been in 
operation since its appointment early in 1948. 

The accomplishments of the present Committee are properly cred- 
ited to the several engineering committees identified with the projects 
briefly described in the balance of this report. 

PRESENTED: As of December 31, 1949. 




A Subcommittee on Cutting and Perforating Raw Stock had been in 
existence since November, 1945, under the chairmanship of Dr. E. 
K. Carver. Since that date the subcommittee has completed several 
assignments, but it was apparent that much standardization of film 
dimensions remained to be done. Accordingly, the Standards Com- 
mittee recommended to the Engineering Vice-President the formation 
of an engineering committee to deal with such problems. In this par- 
ticular field the Standards Committee has acted on the following: 

1. Reaffirmed 2 the American Standard for cutting and perforat- 
ing 35-mm negative raw stock (Z22.34-1944). 

2. Recommended for adoption as standards by ASA two of three 
proposed 32-mm cutting and perforating standards which had been 
published 3 for a period of trial and criticism in the JOURNAL. Not ap- 
proved was the third proposed standard, related to cutting and per- 
forating 32-mm on 35-mm motion picture negative raw stock. It 
was tabled because several members were opposed to having nitrate 
stock perforated in this fashion because of the possibility of its being 
slit to 16-mm and used on conventional 16-mm projection equipment. 

3. Referred back to the Film Dimensions Committee the proposed 
American Standard, "35-mm Motion Picture Combination Positive- 
Negative Raw Stock, Z22.1" which had been published 4 for a period 
of trial and criticism in the JOURNAL. Final action on this proposal 
has been deferred pending the outcome of additional tests of the suit- 
ability of the proposed perforations. 


The subject of standards relating to sprockets continues to be the 
most difficult problem in this category confronting the Committee. 
The entire Committee is in agreement as to the merits of the design 
practice described by Chandler, Lyman, and Martin 5 to insure good 
performance with film. The Committee, however, cannot agree that 
such a practice can properly become an American Standard. Accord- 
ingly, the Committee has asked for and obtained approval from the 
Board to publish an abridgement of the Chandler, Lyman and Martin 
paper in a format suitable for inclusion in the Standards binder of the 
Society as an SMPE recommendation on sprocket design. This 
abridgement is scheduled for publication in an early issue of the JOUR- 

104 FRANK E. CARLSON January 

Concurrently, the Standards Committee has recommended the 
withdrawal of the following two standards relating to this subject 
which are now obsolete: (1) American Standard for 8-mm Motion 
Picture Film Eight Tooth Projector Sprockets (Z22.18-1941); and 
(2) American Standard for 16-mm Motion Picture Film Projector 
Sprockets (Z22.6 1941). 

A subcommittee of the Standards Committee and, later, the 16-mm 
and 8-mm Motion Pictures Committee has been in the process of re- 
viewing and revising the six present standards relating to 16-mm and 
8-mm picture apertures. The Standards Committee has approved 
four proposed standards 6 to take the place of the present six standards 
and has recommended their adoption as standards by ASA and the re- 
scinding of the two superfluous standards Z22. 13-1941 and Z22.14- 

The 16-mm and 8-mm Motion Pictures Committee has also com- 
pleted work on three new standards which the Standards Committee 
has approved for a period of trial and criticism. These three proposed 
standards have been published 7 in the JOURNAL and relate to: 

1. Mounting threads and flange focal distances for lenses on 16- 
mm and 8-mm motion picture cameras. 

2. A base point for focusing scales on 16-mm and 8-mm motion 
picture cameras. 

3. Winding of 16-mm sound motion picture film. 

The first two proposed standards are intended to replace war stand- 
ards developed under the war procedures of the ASA for use of the 
Armed Services. The third proposal is intended to formalize a prac- 
tice in use by film manufacturers for a number of years and which the 
SMPE, in 1941, adopted as a recommended practice. 

A new standard relating to splices for 16-mm motion picture films 
will soon be published in the JOURNAL for a period of trial and criti- 
cism. This proposed standard, if adopted, will replace the two exist- 
ing standards, Z22.24-1941 and Z22.25-1941, relating to silent and 
sound films respectively and is intended to apply only to projection 
films. The Standards Committee feels that a separate standard for 
negative films is also needed and the 16-mm and 8-mm Motion Pic- 
tures Committee has been asked to consider this matter. 

The 16-mm and 8-mm Motion Pictures Committee has also pro- 
posed a revision of Z22. 11-1941 relating to 16-mm motion picture 
projection reels. The Standards Committee is, at the present time, 


reviewing this proposal and may approve it for publication in the 
JOURNAL for a period of trial and criticism. 


Since there is no engineering committee of the Society that is pre- 
pared to take over this project, it has continued to be handled by a 
subcommittee of the Standards Committee under the able chairman- 
ship of Mr. Kingslake. A report of Mr. Kingslake's subcommittee 
has been recently published 8 in the JOURNAL and the subcommittee 
has been requested to prepare this material in a form suitable for con- 
sideration as a proposed standard. 


It is obviously important that the Standards Committee not only 
encourage the processing of new standards beneficial to the industry 
but also recommend the withdrawal of old standards which have 
become obsolete or which, by their continued existence, impede prog- 
ress. Accordingly, the following standards have been reviewed by 
engineering committees, and, as a result of their findings, the Stand- 
ards Committee has recommended withdrawal: (1) American Rec- 
ommended Practice for Motion Picture Film Sensitometry (Z22.26- 
1941) ; and (2) American Recommended Practice for Motion Picture 
Engineering Nomenclature (Z22.30-1941). 


(1) "Report of the Standards Committee," Jour. SMPE, vol. 51, pp. 230-241; 
September, 1948. 

(2) "Standards recommendations," Jour. SMPE, vol. 52, p. 358; March, 

(3) "Three proposed American Standards," Jour. SMPE, vol. 52, pp. 223-230; 
February, 1949. 

(4) "Proposed American Standard," Jour. SMPE, vol. 52, pp. 447-452; April, 

(5) "Proposals for 16-mm and 8-mm sprocket standards," Jour. SMPE, vol. 48, 
p. 483; June, 1947. 

(6) "Proposed standards for 16-mm and 8-mm picture apertures," Jour. SMPE, 
vol. 52, pp. 337-348; March, 1949. 

(7) "Proposed American Standards 16-mm and 8-mm," Jour. SMPE, vol. 53, 
pp. 293-300; September, 1949. 

(8) "Report of lens calibration subcommittee," Jour. SMPE, vol. 53, pp. 368- 
378; October, 1949. 

New American Standards 

THE TWO TEST FILM STANDARDS that appear on the following 
pages were proposed initially and developed by the Motion Pic- 
ture Research Council. After being subjected to careful study under 
the normal procedure of American Standards Association Sectional 
Committee on Motion Pictures Z22, they were approved for publica- 
tion and have now been added to the list of available standards. 

An up-to-date index listing these and several other changes made 
during the past year has just been prepared and copies are available 
to all who use motion picture standards for reference. 

Within the last two or three weeks copies were sent free of charge to 
all whose names are on the Standards Mailing List. If you have one 
of the Society's ' 'American Standards Binders" and have not yet 
received the green index dated January 1, 1950, send your name and 
address to Boyce Nemec, Executive Secretary, and your index will 
be sent by return mail. 

The addresses of all who inquire will be placed on the Binder List 
to be notified each time new or revised standards are approved for 

A number of recent Proposals for Standardization and all of the 
American Standards shown in the index have been printed or reported 
upon in one or more of the following issues of the JOURNAL : 

1946 April page 284 1949 February page 223 
September 258 March 337 

1947 August 171 April 434 
December 547 & 447 

1948 March 280 August 211 
November 534 September 293 



American Standard 

Sound Focusing Test Film for 

35-Millimeter Motion Picture Sound Reproducers 

(Service Type) 

Krf. V. S. Pal. Off. 


UDC 778.5 

1. Scope and Purpose 

1.1 This standard describes a film which may be used for focusing the optical 
systems in 35-millimeter motion picture sound reproducers. The recorded fre- 
quency shall be suitable for use in the routine maintenance and servicing of 
the equipment. 

2. Test Film 

2.1 The film shall be a print from an original negative and shall contain a 
7000-cycle, sinusoidal, variable-area or variable-density track recorded at 1 
decibel below 100-percent modulation. The variation in power output level 
from the film shall be not more than 0.25 decibel. 

2.2 The sound track shall comply with American Standard Sound Record 
and Scanned Area, Z22.40-1946, and the film stock used shall be cut and 
perforated in accordance with American Standard Cutting and Perforating 
Dimensions for 35-Millimeter Motion Picture Raw Stock, Z22.36-1947, or any 
subsequent revision thereof. 

NOTE: A test film in accordance with this standard is available from the Motion Picture 
Research Council or the Society of Motion Picture Engineers. 

Approved February 4, 1949, by the American Standards Association, Incorporated. 
Sponsor: Society of Motion Picture Engineers. 

Copyright, 1949, by the American Standards Assn., Inc. Reprinted by permission of the copyright holder. 



American Standard 

Buzz-Track Test Film for 
35-Millimeter Motion Picture Sound Reproducers 

Rtg. V. S. Pat. Of. 


'UDC 778.534 4 

1. Scope and Purpose 

1.1 This specification describes a film which may be used for checking the 
lateral scanning slit placement of 35-millimeter motion picture sound repro- 

2. Test Film 

2*1 The test film shall be a direct positive recording or a print from an 
originally-recorded negative and shall contain 300-cycle and 1000-cycle 
square-wave tracks on either side of the central exposed strip as shown in 
Fig. 1. 

0.269 IN. MAX 
0.287 IN. MIN 
7.34 MM MAX 
7.29 MM MIN 


0.201 IN. MAX 
0.200 IN. MIN 
5.10 MM MAX 
5.08 MM MIN 

Fig. 1 

2.2 The central exposed strip and the exposed portion of the-two signal tracks 
shall have a minimum density of 1 .4 and a maximum density of 2.0. 

2.3 The film stock used shall be cut and perforated in accordance with the 
American Standard Cutting and Perforating Dimensions for 35-Millimeter 
Motion Picture Positive Raw Stock, Z22.36-1947, or the latest revision thereof 
approved by the American Standards Association, Incorporated. 

2.4 The film stock used shall have a shrinkage of not more than 0.50 percent. 

NOTE: A test film in accordance with this standard is available from the Motion Picture 
Research Council, or the Society of Motion Picture Engineers. 

Approved May 23, 1949, by the American Standards Association, Incorporated. 
Sponsor: Society of Motion Picture Engineers. 

Copyright, 1949, by the American Standards Assn.^lnc. Reprinted by permission of the copyright holder. 

New Officers of the Society 

Members who were elected during tlio annual Society elections in 
1949 took office on the first of January. So that Society members who 
have recently joined may become better acquainted with these men 
who are responsible for managing the affairs of the Society, here they 

Engineering Vice-President 

Fred T. Bowditch has been closely associated with Society adminis- 
tration for many years, having served several terms as a member of 
the Board of Governors. This year he vacates the second half of a 
two-year Governorship, in order to become Engineering Vice-Presi- 
dent for 1950 and 1951 ; and, so that he may devote a major part of his 
time to the important engineering committee work of the Society, he 
has also resigned from the Chairmanship of Committee Z22, the Sec- 
tionarCommittee on Motion Pictures of the American Standards As- 
sociation. Since last October Mr. Bowditch, John A. Maurer, outgo- 
ing Engineering Vice-President and Bill Deacy, Society Staff Engi- 
neer, have been reviewing committee appointments and projects, with 
an eye toward the easy transfer of responsibilities, as well as the grad- 
ual expansion of the work of our many technical committees. 

Mr. Bowditch is Director of Illuminating Carbon Research for the 
National Carbon Company, Box 6087, Cleveland 1. 

Financial Vice-President 

Ralph B. Austrian moves from the office of Treasurer which he held in 
1948 and 1949 to become Financial Vice-President for 1950-1951. 
He replaces David B. Joy, who retires from the Board of Governors 
this year after a number of years of very active service to the Society. 
Long associated with the sound recording and theater equipment 
part of the motion picture industry, Mr. Austrian has more recently 
been active in the field of television. He is currently doing consulting 
work in television broadcasting and also in theater television. He 
may be reached at 25 West 54th St., New York 19. 


Frank E. Cahill becomes Treasurer of the Society for 1950 and 1951, 
replacing Mr. Austrian. Having been a member of the Society for 
nearly twenty years, with many years of service on the Board of Gov- 


ernors and on several engineering committees, Mr. Cahill is well 
known to the Society and to the motion picture industry. During the 
recent war, he was Executive Officer of Army Pictorial Service and at 
the end of the war, he reassumed his duties as Technical Director of 
Warner Bros. Circuit Management Corp., 321 West 44th St., New 
York 18. 


Lorin D. Grignon begins his first term on the Board of Governors after 
a number of active years with the Pacific Coast Section of the Society. 
"Flicker in Motion Pictures" has long been a topic of major interest 
for Mr. Grignon and more recently he has been active in television 
film and theater television matters on the West Coast. He is an engi- 
neer in the Studio Sound Department at Twentieth Century-Fox 
Films, Inc., Box 900, Beverly Hills, Calif. 

Paul J. Larsen was elected last year to his third successive term as 
a Governor of the Society. His interest in theater television continues 
although he is currently occupied with other matters. He may be 
reached at 508 S. Tulane St., Albuquerque, N.M. 

William H. Rivers has served two one-year terms as Chairman of 
the Atlantic Coast Section of the Society, which made him an ex- 
officio member of the Board for 1948 and 1949. As a result, members 
in and near New York City are well acquainted with him. He now 
begins his first two-year term as an elected Governor and may be 
reached at Eastman Kodak Company, 342 Madison Ave., New York 

Edward S. Seeley was a Manager of the Atlantic Coast Section for a 
two-year term which ended in December, 1949, and he now becomes a 
Governor for 1950 and 1951. Having been Chief Engineer of Altec 
Service Corp. for a number of years, Mr. Seeley is well known in the 
audio-engineering and theater equipment fields. His office is at Altec 
Service Corp., 161 Sixth Ave., New York 13. 

R. T. Van Niman has served two successive terms as Chairman of 
the Central Section, for 1948 and 1949. In 1948 he was elected to two 
Society offices : Central Section Chairman for 1949 and Society Gov- 
ernor for 1949 and 1950. He chose to resign the Governorship in or- 
der to devote more time to the work of the Section. He now retires 
from the Section Chairmanship and has once again been elected as a 
Governor, this time for 1950 and 1951. As a member of the sound 
committee for several years he was Chairman of the Subcommittee 
on Phototubes. He is a sound engineer for Motiograph, Inc., and may 
be reached at 4501 Washington St., Chicago 24. 

John P. Livadary was elected in 1949 to fill a one-year West Coast 
Governorship vacancy. The unfilled position was left vacant by the 
resignation of S. P. Solow who in 1948 was elected to two Society of- 
fices and chose to retain the Chairmanship of the Pacific Coast Sec- 
tion. Mr. Livadary is a member and friend of the Society of long 
standing who now begins his first term on the Board. Having been 
Sound Director of Columbia Pictures for many years, he is active in the 
work of the Motion Picture Research Council and of our own sound 
committee, and may be reached at 4034 Cromwell Ave., Los Angeles 

Section Officers 

The Atlantic Coast Section Chairman for 1950 is Edward Schmidt, 
Technical Representative for Reeves Sound Studio, 304 East 44th 
St., New York 17. The Secretary-Treasurer is Harry Milholland, 
head of .the Tele-Transcription Department at Allen B. Du Mont 
Laboratories, Inc., 515 Madison Ave., New York 22. 

Central Section Chairman is George W. Colburn, President of 
George W. Colburn Laboratories, 164 No. Wacker Drive, Chicago 6. 
The Secretary-Treasurer is C. E. Heppberger, Lighting Carbon Tech- 
nical Specialist, National Carbon Company, Inc., 230 No. Michigan 
Ave., Chicago 1. 

Pacific Coast Chairman is Charles R. Daily, Optical Engineer, 
Paramount Pictures, Inc., 5451 Marathon St., Hollywood 38,. Calif. 
The Secretary-Treasurer is Vaughn C. Shaner, Technical Service, 
Eastman Kodak Company, 6706 Santa Monica Blvd., Hollywood 38, 

Engineering Committees 

In addition to appointing new chairmen for several of the Society's 
eighteen engineering committees, F. T. Bowditch, newly elected En- 
gineering Vice-President, and Bill Deacy, Society Staff Engineer, have 
just completed an impressive schedule of engineering committee proj- 
ects for 1950. Most of them are a continuation of work that was be- 
gun during 1949 or earlier, but several of the projects are entirely new. 
One example concerns a special leader for 16-mm television films that 
would replace the conventional "academy" leader that is generally 
not favored by television film users. One of the requirements of the 


new leader is that it give an accurate "on the air" cue for both the 
television projector operator and the program director. This is a se- 
rious matter to television broadcasters since film must be cued to start 
on the second with no such liberal tolerances as are standard practice 
in theaters. 

Another of the new projects, and one that was suggested by the 
Technical Editor of a motion picture trade magazine, has to do with 
possible standards or recommendations for air conditioning of theater 

So that members may keep posted throughout the year on the cur- 
rent status of these and other projects, each issue of the JOURNAL for 
1950 will review some of the work of several engineering committees. 
This is something new and Bill Deacy would appreciate comments on 
this method of reporting. 

Frank E. Carlson of General Electric, Nela Park, Cleveland, has 
just been appointed Chairman of the Standards Committee for his 
second consecutive two-year term. In this issue of the JOURNAL he re- 
ports on changes made in the organization of the Standards Commit- 
tee when he became chairman early in 1948, and on the accomplish- 
ments of the committee over the past two years. During that time 
the Standards Committee has served primarily as a policy-making 
body, supervising the Society's standardization activities and advis- 
ing the Engineering Vice-President. A careful study of Mr. Carlson's 
report will give a good working knowledge of the way in which stand- 
ards are developed; and his report should be of interest to members 
who either use these American Standards or participate in some phase 
of the Society's committee work. 

The final result of almost any engineering project has in the past 
been either a specific formal standards proposal or a detailed commit- 
tee report which, though generally not as concise as a standard, has 
served to document a particular chapter or phase of our engineering- 
history. For some time we have needed an "in-between" type of docu- 
ment, less formal than a standard but more specific than the custom- 
ary committee report. It should be a way of presenting committee 
recommendations as a series of reference publications on subjects that 
do not lend themselves readily to standardization under rigid ASA 
procedures. To fill this gap, the Society's Board of Governors re- 
cently approved publication of "Society Recommendations." A de- 
scription of this new type of publication is part of the report of the 
President for 1949 that appears in this issue. 


1950 Nominations 

All Active, Fellow and Honorary members may recommend candi- 
dates for the ten vacancies on the Board of Governors which will occur 
on December 31, 1950. Suggestions should be mailed early so that 
they will certainly be in the hands of the Nominating Committee prior 
to May 1, 1950. They may be addressed to the Chairman or to any 
of the members of the Committee : 

D. E. Hyndman, Chairman 
Room 626, 342 Madison Ave., New York 17. 

Herbert Barnett G. R. Giroux 

General Precision Equipment Corp. Technicolor Motion Picture Corp. 

63 Bedford Road, Pleasantville, N.Y. 631 1 Romaine St., Los Angeles 28, Calif. 

F. T. Bowditch A N Goldsmith 

Research Laboratories 597 Fifth Ave., New York 17. 

National Carbon Company 

Box 6087, Cleveland 1. T. T. Goldsmith 

F E Cahill Jr Allen B. Du Mont Laboratories 

Warner Bros. Pictures, Inc. 2 Main Ave - Passaic, N.J. 

321 West 44th St., New York 20. K F M n 

R. E. Farnham Western Electric Company 

General Electric Company 6601 Romaine St., 

Nela Park, Cleveland 12. Hollywood 38, Calif. 

The Board members whose terms expire at the end of this year are : 

President, E. I. Sponable Governor (East), Herbert Barnett 

Executive Vice-President, Peter Mole Governor (East), (vacant) 

Editorial Vice-President, C. R. Keith Governor (West), K. F. Morgan 

Secretary, R. M. Corbin Governor (West), J. P. Liyadary 

ConventionVice-Pres.,W.C.Kunzmann Governor (West), N. L. Simmons 

Society Awards for 1950 

Each year the Society considers candidates for five awards on the 
basis of the qualifications outlined briefly here. Further details con- 
cerning these awards are published in the April issue of the JOURNAL 
each year for the information of members who may not be familiar 
with them. Suggestions or questions concerning these matters may 
be addressed to the chairman of any of the award committees or to 
the Executive Secretary at Society Headquarters in New York. 

Fellow Award 

Members in the Active grade who by their " . . . proficiency 
and contributions have attained outstanding rank among engineers or 
executives of the motion picture industry" may be proposed and con- 


sidered as possible award nominees by the Fellow Award Committee. 
Such proposals will be received only from present Fellows of the So- 
ciety and should be addressed to Loren L. Ryder, Committee Chair- 
man and Past-President of the Society. His address is: Paramount 
Pictures, Inc., 5451 Marathon St., Hollywood 38, Calif. 


The Honorary Membership Award is a distinction given to pioneers 
who have contributed inventions of basic importance to the industry 
or whose contributions have made possible better production, admin- 
istration or utilization of motion pictures. Recommendations for the 
Honorary Membership Award may be submitted by any member of 
the Society and must be endorsed by at least five Fellows, who are re- 
quired to set forth in writing their knowledge of the accomplishments 
which appear to justify presentation of the Award. Such recom- 
mendations must be addressed to the Honorary Membership Commit- 
tee Chairman, Gordon A. Chambers, Motion Picture Film Dept., 
Eastman Kodak Company, 343 State St., Rochester 4, N.Y. 

Journal Award 

The Journal Award is presented annually at the Fall Convention of 
the Society to the author of the most outstanding paper originally pub- 
lished in the JOURNAL of the Society during the preceding calendar 
year. Technical merit, originality and excellence of presentation are 
three major considerations. The authors of papers of nearly equiva- 
lent merit often receive Honorable Mention. The Journal Award 
Committee, appointed by the President, is now under the Chairman- 
ship of Dr. C. R. Daily, who will shortly be receiving from members of 
his Committee, their recommendations for the most outstanding pa- 
per for 1949. His address is : 5451 Marathon St., Hollywood 38, Calif. 

Samuel L. Warner Memorial Award 

Each year the President appoints a Samuel L. Warner Memorial 
Award Committee to consider candidates for the Award. Preference 
is given to inventions or developments occurring in the last five years, 
and also to inventions or developments likely to have the widest and 
most beneficial effect on the quality of reproduced sound and pic- 
ture. The Award is made to an individual and may be based upon his 
contributions of the basic idea involved in the particular development 
being considered and also on his contributions toward the practical 
working out of the idea. The purpose of the Award is to encourage 
the development of new and improved methods or apparatus designed 
for sound on film motion pictures, including any step in the process. 

The present Chairman of the Committee is W. V. Wolfe, Motion Pic- 
ture Research Council, 1421 North Western Ave., Hollywood 27, 

Progress Medal Award 

Written recommendations for candidates for the Progress Medal 
Award may be submitted by any member of the Society, giving in de- 
tail the accomplishments which appear to justify consideration. 
Qualifications should include invention, research, or development 
which has resulted in a significant advance in the development of mo- 
tion picture technology and should be seconded in writing by any two 
Fellows or Active members of the Society, after which the recom- 
mendations must be filed with the Chairman of the Committee. For 
1950, the Chairman is Dr. J. G. Frayne, Westrex Corp., 6601 Romaine 
St., Los Angeles 38, Calif. 

Society Announcements 

Membership Directory 1950 

The Society's Membership Directory for 1950 is scheduled for early publication, 
so members are urged to reply promptly to the questionnaire-envelope enclosed 
with the 1950 membership dues statement. Since the Directory is published 
every two years, addresses and company affiliations must be corrected how, or 
stand uncorrected till 1952. Please send yours in now if it has not already been 

Section Meetings 

The first regular meeting of the Atlantic Coast Section in 1950 is scheduled for 
7:30 P.M., January 18, at the RKO Pathe Studios, 105 East 106th St., New York 
City. Ed Schmidt, Section Chairman, reports that Drs. E. B. Jennings and A. B. 
Weiss will describe and demonstrate the new du Pont release positive color film. 
This is to be a somewhat condensed version of several du Pont papers on the new 
Type 275 film that were presented during the 66th Convention in Hollywood last 

Attendance should be excellent since this will be the first opportunity that many 
East Coast members of the Society have had to look over this much .discussed 
new film. 

The Central Section meets at 8:00 P.M., January 12, in the small auditorium of 
the Western Society of Engineers, 84 E. Randolph St., Chicago 1. Current trends 
in film distribution will be dsucussed by Thomas McConnell, Chicago attorney and 
authority on that phase of the motion picture industry. 

This will be a double feature meeting, with a second paper presented by George 
W. Colburn, describing a Double System 16-mm Projector. Educational and 


advertising film producers who work exclusively with 16-mm films should be 
greatly interested, since Mr. Colburn will describe the modification of a conven- 
tional 16-mm projector which enables it to reproduce separate films for picture and 
sound track. 

67th Convention 

Plans are well underway for the 67th Society Convention to be held at the 
Drake Hotel in Chicago, April 24-28th. Members and their guests who plan to 
attend will be assured of adequate hotel accommodations since many rooms have 
been set aside by the hotel. These and other arrangements have already been 
made by Bill Kunzmann, our perennial Convention Vice-President. The Papers 
Program is already beginning to take form and the Papers Committee is hard at 

An advance schedule of convention events, with hotel room reservation forms, 
will be mailed to all members about March 1. The Papers Committee expects 
to mail, about the last week in March, the Tentative Program listing all papers 
then scheduled. Authors who expect to present papers should ask the nearest 
Papers Committee Vice-Chairman for their Author's Forms. Copies of the 
Author's Form, together with a 100-word abstract of each paper should be sent to 
R. T. Van Niman in Chicago, so that the paper can be scheduled in time for the 
Tentative Program. 


N. L. Simmons, Chairman E. S. Seeley 

6706 Santa Monica Blvd., Altec Service Corp., 

Hollywood, Calif. 161 Sixth Ave., New York 13, N.Y. 

Vice-Chairmen: R. T. Van Niman 

J. E. Aiken 4501 Washington Blvd., 

116 N. Galveston St., Arlington, Va. Chicago 24, 111. 

L. D. Grignon H. S. Walker 

20th Century-Fox Films Corp., 1620 Notre Dame St., W., 

Beverly Hills, Calif. Montreal, Que., Canada 

The Editor 

This month's masthead shows for the first time our new Editor Vic Allen, 
who joined us late in November. In previous experience as Production Manager 
for Interscience Publishers, he put into production Bill Offenhauser's book, 16-Mm 
Sound Motion Pictures. He is well acquainted with the Society's printer, Mack 
Printing Company, in Easton, Pennsylvania, having handled several thousand 
pages printed yearly by Mack for Interscience. Before joining Interscience, he 
was Managing Editor of the Journal of the American Water Works Association. 
He began his technical editing on The Foundry magazine of Penton Publishing 
Company in Cleveland while a "co-operative" student from Antioch College. 
He has contributed to the publishing and printing field by active membership in 
the American Institute of Graphic Arts and by articles on such subjects as: vari- 
ous methods of typesetting tabular matter; two-column format for technical 
books; and methods of drafting technical illustrations. 


Book Reviews 

Acoustic Measurements, by Leo L. Beranek 

Published (1949) by John Wiley & Sons, 440 Fourth Ave., New York 16. 
896 pp. + 17 pp. index + VII pp. " 519 illus. 6 X 8'A in. Price, $7.00. 

This much needed book is a comprehensive collection of techniques and of 
tables of constants which the acoustic engineer requires for measurements and 
calculations. Of interest is a brief history of acoustic measuring instruments; 
and of reference value is a glossary of terminology. Dr. Beranek presents the 
solutions of the sound wave equations in various forms, with complete data on the 
velocity and attenuation in a great variety of media including effects of wind, 
jungle growth, etc.; and he also gives the experimental and calculated diffraction 
effects due to variously shaped bodies placed in the path of a plane. 

Then follows an excellent treatise on techniques of calibrating microphones as 
standards for measuring sound pressures with particular emphasis on the reci- 
procity method. The methods generally used for measuring frequency are clearly 
presented with some good photographs of commercial instruments available 

The chapter on the principles of calibrating pure tone audiometers is timely 
because of current efforts to develop specifications for a standard audiometer. 
The author discusses various types of meters for measuring quantities related 
to sound intensity such as peak meters, V. U. meters, level recorders, RMS meters, 
and also meters for analyzing transient and steady sounds into various sorts of 
components. The basic tests for the efficiency of communication systems to 
transmit speech are itemized and described, such as methods of measuring fre- 
quency and nonlinear response characteristics, repetition counts, syllable, word, 
and sentence articulation tests, and threshold measurements of received speech. 

Methods are given in detail for testing the three basic elements of a communica- 
tion system: the microphone, the line (including amplifier), and the headphone 
or loudspeaker. In each case the author has outlined the method of testing the 
frequency response, the power efficiency, the impedance, the nonlinear distortion, 
and overload capacity. With the discussion on loudspeakers there is a useful 
set of curves for determining power rating to give satisfactory loudness of speech 
or music in rooms of any size and treatment. There is one chapter on real voice 
testing methods of determining response characteristics of microphones and ear- 
phones. It is shown how these principles can be applied in a convenient form for 
testing the important characteristics of hearing aids. Methods of making articu- 
lation tests are outlined together with lists of syllables, of words, and of sentences, 
including the P. B. and Spondee tests. 

The last three chapters are devoted to measurements of the acoustic properties 
of rooms, including the various methods of measuring the absorption properties 
of materials for treatment of such rooms. References throughout the book are 
numerous and should permit a student to pursue very satisfactorily any special 
phase. Many engineers will be grateful to Dr. Beranek for bringing together in 
such a convenient form so much technical information bearing on acoustic meas- 


(Columbia University) 

5 Westminster Rd. 

Summit, N.J. 


Painting with Light, by John Alton 

Published (1949) by Macmillan, 60 Fifth Ave., New York 11. 191 pp. + 
XIV pp. 292 figs. 7 3 A X 10 ! A in. Price, $6.00. 

It has been said that the mechanical techniques of an art should be learned, then 
forgotten. More properly stated, they should become an unconscious part of 
the work of an artist. The author of Painting with Light frankly describes the 
techniques and devices he uses to obtain his effects. He does not assume that a 
novice may become a Director of Photography by reading his book and studying 
the copious layouts and illustrations, but he does describe his work with a straight- 
forwardness that is refreshing as well as instructive. 

While the book may seem to be an oversimplification of a very complicated art 
form to his brother workers in motion picture photography, it will serve as a means 
of conveying some of the problems of the cinematographer to many other depart- 
ments of the industry, as well as to the associated organizations which design and 
manufacture equipment and materials for the industry. 

The motion picture industry has a language of its own for describing the various 
workers and accoutrements used in motion picture set lighting and the book acts 
as an interpreter for the uninitiated. 

The various types of lamps and lighting control equipment are described and 
illustrated. Lamp placement and manipulation are explained and illustrated with 
layouts as well as with photographs of the end results. 

Chapters cover both indoor and outdoor photography, the close-up, long-shot, 
process work, and miniatures. For the most part, the author does not deviate 
from his subject and, while some of his techniques such as the "testlight" are not 
universal, he has spared neither time nor expense to cover the subject as com- 
pletely as possible in so far as black and white photography is concerned, from his 
own viewpoint and within the covers of one book. 


Director of Photography 

139 a /2 So. Doheny Drive 

Los Angeles 48, Calif. 

Feininger on Photography, by Andreas Feininger 

Published (1949) by Ziff-Davis, 350 Fifth Ave., New York 1. 409 pp., 360 
illus. (approx.), 50 charts (approx.). 8 x /2 X 11 in. Price, $15.00. 

It is rare that a reviewer for a technical journal can go all out in praise of a basic 
book on photography with no fear that he is exposing himself to criticism. But 
this book is one that even the astute technician will consider well done for the pur- 
pose intended. 

Mr. Feininger has put down for the amateur and professional still photographer 
what his 20 years of experience have shown to be essential to good picture making. 
Although he de-emphasizes matters of a strictly technical nature, he advocates 
systematic working methods based on fact and not folly. In reading the text one 
cannot help but be impressed by the unusually clear and, for the most part, ac- 
curate, insight into technical matters that the author has gained from his experi- 
ence. It even appears that he may have studied the better technical literature 
to a greater extent than he recommends. 

The book covers the subject thoroughly in 16 chapters. Little time is wasted 
anywhere in getting to the point, for the author is no believer in secrets or mystery 
in the photographic process; but he does stress that technical knowledge alone 
will never make a good photographer. This requires MM "eye" tW pictures which 
you either do or do not have. This attitude explains \\hy Part I on technique 
consists of but seven chapters, whereas Part II on the art of nuking :i p holo- 
graph contains nine chapters. 

Little would be gained by giving the usual list of chapter hemlines. The 
real value in the book will be found only in reading it page by page. It is highly 
recommended, especially to the motion picture engineer who seldom takes pic- 
tures, but who now and then gets the yen to "show up those guys at Life." 
Here's your chance, for Feininger, one of Life's most famous and able photogra- 
phers, has left very little unsaid. 


Pavelle Color Inc. 

533 West 57th St. 

New York 19 

The Complete Projectionist, by R. Howard Cricks 

Published (1949) by Odhams Press, 6 Catherine St., London, W.C. 2. 335 
pp. + 38 pp., including 14-page index. 209 illus. + tipped-in folded insert. 
5 X 7 1 / 2 in. Price, 10/ - post free. 

This work, obviously intended as a handbook for British projectionists, covers 
projection from every angle. By virtue of the fact that the author does cover the 
entire scope of the craft in 335 pp., comprehensive description of any single phase 
of projection is necessarily lacking. The numerous tables, charts, and illustra- 
tions are extremely well presented and will prove valuable to any projectionist or 
projection engineer. 

Mr. Cricks' inclusion of television and several experimental developments will 
prove interesting to the craft as a whole. His chapters on projection practices 
in other than theater locations (process projection, 16-mm projection, etc.) are 
more descriptive of the job than of the technical operation of the equipment. 

Despite the fact that data on equipment, rules, and regulations are necessarily 
limited to the British, the work will prove an informative addition to the library 
of any member of the craft. 


Supervisor of Projection 

Metro Goldwyn Mayer Studios 

Culver City, Calif. 

EMPLOYMENT SERVICE Cameraman-Director, currently em- 

wAivrTiyn ployed by internationally known pro- 

WANTED ducer, desires greater production oppor- 

tunities. Fully experienced 35- and 
16-mm, color, b & w; working knowl- 
edge editing, sound, and laboratory 
problems; administrative experience. 
Top references and record of experience 
available. Write P.O. Box 5402, Chi- 


Current Literature 

THE EDITORS present for convenient reference a list of articles dealing with 
subjects cognate to motion picture engineering published in a number of 
selected journals. Photostatic or microfilm copies of articles in magazines that 
are available may be obtained from The Library of Congress, Washington, D.C., 
or from the New York Public Library, New York, N.Y., at prevailing rates. 

American Cinematographer 

vol. 30, no. 9, September, 1949 

The Garutzo Lens in Motion Picture 
Photography (p. 320) R. M. NEW- 

Eclair Camerette Makes U.S. Debut 
(p. 321) B. BERG 

Source Lighting (p. 324) C. LORING 

Teaching Speech with 16-mm Movies 
(p. 330) R. W. STANMYRE 

vol. 30, no. 10, October, 1949 
They Do it with Infra-Red! (p. 360) 

The Magic of Montage (p. 361) HERB 

Balancing Television Camera Tubes (p. 


vol. 30, no. 11, November, 

Lighting Translucent Backings (p. 398) 

Signal System (p. 402) L. ALLEN 

vol. 30, no. 12, December, 


New Speed for Films (p. 440) L. ALLEN 

A 16-mm Sound Camera for the Home 

Movie Maker (p. 444) G. B. LEWIS 

British Kinematography 

vol. 15, no. 2, August, 1949 
Developments in Magnetic Sound-on- 
Film Recording 

Pt. I, Magnetic Sound and the Film 
(p. 37) O. K. KOLB 

Pt. II, Magnetic Recording in Film 
Production (p. 47) N. LEEVERS 

Pt. Ill, Performance Data of Mag- 
netic Coatings (p. 50) A. TTJTCH- 

Demonstration of Sub-Standard Kine- 
matograph Equipment 

Danson Projector D23 (p. 54) W. 

Debrie D16 Projector with Arc Lamp 
(p. 55) T. A. BARTLETT 

Long-Running Projector with Mer- 
cury Lamp (p. 55) P. J. ORAM 

M. R. Type 356 Cine-Flash (p. 56) 

"Brook" Continuous Projector (p. 
57) H. S. HIND 


vol. 15, no. 3, September, 1949 
Closed Sequence Control Systems (p. 

75) E. B. PEARSON and A. PORTER 
The Application of Music to Films (p. 


vol. 15, no. 4, October, 1949 
Thirty Years of British Film Produc- 
tion (p. 105) M. BALCON 
Current Practice in 16-mm Sound Print- 
ing (p. 116) M. V. HOARE 

vol. 15, no. 5, November, 1949 
Independent Frame Film Production: 

I. Notes on a Production Tech- 
nique (p. 141) K. BELLMAN 

II. Progress Report on Independ- 
ent Frame (p. 150) D. WILSON 

Screen Illumination with Respect to 
Optics and Arc Characteristics (p. 
155) A. G. DUERDOTH 


vol. 22, no. 12, December, 

New Directions in Color Television (p. 


Dot Systems of Color Television, Pt. 1 
(p. 88) W. BOOTHROYD 

International Photographer 

vol. 21, no. 9, September, 1949 
Light Can be Packaged (p. 20), A. 

International Projectionist 

vol. 24, no. 9, September, 1949 

Lens and Film Factors Affecting Focus, 
Pt. 2 (p. 7) R. A. MITCHELL 

The 'Arcon' Projection Arc Monitor 
(p. 10) V. G. MATHISEN 

New Series of Lenses for 16-mm Pro- 
fessional Projection (p. 16) A. E. 

vol. 24, no. 11, November, 

The 35-mm Projection Positive Film 

(p. 8) R. A. MITCHELL 
Theatre Television: What, How and 
When (p. 12) J. E. McCoY and H. 

Film Fire Characteristics (p. 16) R. D. 

New Products 

Further information concerning the material described below can 
be obtained by writing direct to the manufacturers. As in the case 
of technical papers, publication of these news items does not consti- 
tute endorsement of the manufacturer's statements nor of his products. 

The Bell & Howell Design 2709 16-mm Camera is B&H's answer to the growing 
demands of the professional 16-mm field. It is an adaptation of the B&H Design 
2709 Standard Camera (35-mm), with these reported features: (1) a 4-lens turret 
designed to permit the use of all the standard professional lenses; (2) a fixed-pilot- 
pin film movement mechanism similar to the B&H Unit "I"; (3) a 170 adjustable 
shutter with automatic dissolve; and (4) adaptability of stop-motion motor for 
one-, two- or three-frame operation. 

The 200-, 400- and 1000-ft standard B&H (35-mm) magazines may be adapted 
to the 16-mm size by double flanges, rollers and cores. 

Information is available from the Professional Equipment Div., Bell & Howell 
Co., 7100 McCormick Rd., Chicago 45. 


The Movie Master is a new timepiece 
manufactured by Atlas Time Corp., 
2 West 47th St., New York 19. A 
forerunner of the Movie Master timer 
was the stop watch and timer made by 
Moss and Robinson, a subsidiary of 
Atlas Time Corp. The new timer has 
the same general features: three outer 
scales dividing the minute into 90 
parts; three inner scales dividing the 
minute into 36 parts; and an extreme 
inner scale or minute track dividing 
the minute into 60 sec. It denotes in 
the first three minutes which of the 
colored scales of either 35- or 16-mm is 
in use. It has a 5Y 2 -in. dial and stands 
on a table. The price is $12.50. 

This Densitometer has been developed by the Photo volt Corp., 95 Madison Ave., 
New York 16, and is fully described in that firm's Bulletin No. 245. It is a photo- 
electric instrument designed to cover the transmission density range from to 5.0. 
The range from to 2.0 is direct reading, while other ranges may be selected by the 
operator and involve the use of an addition factor. 

The densitometer is provided with a traveling film guide that accommodates 
both 16- and 35-mm film with the sound track always in register with an illumi- 
nated slit 0.020 in. wide. It also "reads" Eastman lib Sensitometer strips and the 
smaller strips from the Eastman Processing Control Sensitometer, replacing more 
elaborate and more expensive equipment which in the past has been used for mak- 
ing these measurements on black-and-white films. 


The Weston Cadet is a new exposure 
meter designed especially for travelers 
and casual photographers who want a 
small, easy-to-use meter. Its list 
price is $21.50. It is made by the 
Weston Electrical Instrument Corp., 
617 Frelinghuysen Ave., Newark 5, 
N.J. Though small enough to fit into 
a vest pocket or purse, it is equipped 
with the Weston instrument movement 
and photronic cell, and is designed to 
give accurate shutter and diaphragm 
settings for all general amateur pho- 
tography with either still or motion pic- 
ture cameras, and for both black-and- 
white and color film. The Weston 
Cadet can be used for measuring inci- 
dent light by using a translucent light 
collector which is pivoted on the back 
of the meter. 

Meetings of Other Societies 

March, 1950 

Institute of Radio Engineers 

National Convention 
Optical Society of America 
Winter Meeting 

March 6 through March 9 
New York, New York 
March 9 through March 11 
New York, New York 

May, 1950 
Armed Forces 

Communications Association 
Annual Meeting 

May 12 

New York City and Long Island City 

May 13 

Fort Monmouth, New Jersey 

June, 1950 

Acoustical Society of America 

June 22-24 

State College, Pennsylvania 

SMPTE Officers and Committees: The names of Society Officers 
and of Committee Chairmen and Members are published annually in 
the April issue of the JOURNAL. Changes and corrections to these 
listings are published in the September JOURNAL. 


The Blue Comet Boom Light has been 
developed by Mole-Richardson Co. of 
Hollywood, Calif., to supply flexible 
illumination in commercial, motion pic- 
ture and television studios. A new 
feature described by the manufacturer 
is the light-weight Blue Zephyr baby 
spot, with attached full-size diffusion 
frame and rotating barn doors, plus 
direct-action focusing with graduated 
scale. The lamp head is interchange- 
able with boom or the included auxil- 
iary stand. To facilitate handling and 
transporting, the stand and boom legs 
are designed to fold flat, and the stand 
is strongly constructed with positive 
locking fittings and 3-in. rubber-tired 
casters with locking feet. The stand is 
reported not to tip even when operated 
at a y fully extended position, and the 
stand and boom arm are designed for 
easy extension, with an air-retainer 
brake in the boom stand to permit 
smooth, quiet adjustment of height. 


By action of the Board of Governors, October 4, 1931, this Honor Roll was 
established for the purpose of perpetuating the names of distinguished pioneers 
who are now deceased. 

Louis Aime Augustin Le Prince 

William Friese- Greene 

Thomas Alva Edison 

George Eastman 

Frederic Eugene Ives 

Jean Acme Le Roy 

C. Francis Jenkins 

Eugene Augustin Lauste 

William Kennedy Laurie Dickson 

Edwin Stanton Porter 
Herman A. DeVry 

Robert W. Paul 
Frank H. Richardson 

Leon Gaumont 
Theodore W. Case 

Edward B. Craft 
Samuel L. Warner 

Louis Lumiere 

Thomas Armat 


Lee de Forest 

A. S. Howell 


Journal of the Society of 

Motion Picture and Television Engineers 


Basic Research for Motion Pictures CLYDE R. KEITH 127 

Noise Considerations in Sound-Recording Transmission Systems 

F. L. HOPPER 129 

Photography in the Rocket-Test Program CARLOS H. ELMER 140 

High-Speed Processing of 35-Mm Pictures . 


The Trend in Drive-in Theaters CHARLES R. UNDERBILL, JR. 161 

A Sturdy, High-Quality 16-Mm Projector 

G. T. LORANCE, F. B. DIBBLE, and H. J. REED 171 

Animar Series of Photographic Lenses 


Color Cinematography in the Mines M. CHARLES LINKO 199 

Television Test Film 209 

Recommendations for 16-Mm and 8-Mm Sprocket Design 219 

Proposed American Standard for 16-Mm Projection Reels 229 

A Restatement of Policy 233 

Board of Governors 234 

67th Semiannual Convention 235 

Society Announcements 236 

Engineering Committees 237 

Section Meetings 240 


16-Mm Sound Motion Pictures, by William H. Offenhauser, Jr 

Reviewed by Lloyd T. Goldsmith 241 

The Recording and Reproduction of Sound, by Oliver Read 

Reviewed by O. B. Gunby 242 

Meetings of Other Societies 242 

New Products 243 

Employment Service 244 


Chairman Editor Chairman 

Board of Editors Papers Committee 

Subscription to nonmembers, $12.50 per annum; to members, $6.25 per annum, included in 
their annual membership dues; single copies, $1.50. Order from the Society's General Office. 
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and 
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture Engineers, 
Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General and Editorial Office, 
342 Madison Ave., New York 17, N. Y. Entered as second-class matter January 15, 1930, 
at the Post Office at Easton, Pa., under the Act of March 3, 1879. 

Copyright, 1950, by the Society of Motion Picture Engineers, Inc. Permission to republish 
material from the JOURNAL, must be obtained in writing from the General Office of the Society. 
Copyright under International Copyright Convention and Pan-American Convention. The 
Society is not responsible for statements of authors or contributors. 

Society of 

Motion Picture and Television Engineers 

342 MADISON AVENUE, NEW YORK 17, N.Y., TEL. Mu 2-2185 

Earl I. Sponable 
460 W. 54 St. 
New York, 19, N.Y. 

Peter Mole 

941 N. Sycamore Ave. 
Hollywood 38, Calif. 



Loren L. Ryder 
5451 Marathon St. 
Hollywood 38, Calif. 

Clyde R. Keith 
120 Broadway 
New York 5, N.Y. 

Robert M. Corbin 
343 State St. 
Rochester 4, NY. 

William C. Kunzmann 
Box 6087 
Cleveland 1, Ohio 

Fred T. Bowditch 
Box 6087 
Cleveland 1, Ohio 


Frank E. Cahill, Jr. 
321 W. 44 St. 
New York 18, N.Y. 



Ralph B. Austrian 

25 W. 54 St. 

New York 19, N.Y. 


Herbert Barnett 
Manville Lane 
Pleasantville, N.Y. 

Kenneth F. Morgan 
6601 Romaine St. 
Los Angeles 38, Calif. 

Norwood L. Simmons 
6706 Santa Monica Blvd. 
Hollywood 38, Calif. 


George W. Colburn 
164 N. Wacker Dr. 
Chicago 6, 111. 

Charles R. Daily 
113 N. Laurel Ave. 
Los Angeles 36, Calif. 

John P. Livadary 
4034 Cromwell Ave. 
Los Angeles, Calif. 

William B. Lodge 
485 Madison Ave. 
New York 22, N.Y. 

Edward Schmidt 
304 E. 44 St. 
New York 17, N.Y. 


Lorin D. Grignon 
1427 Warnall Ave. 
Los Angeles 24, Calif. 

Paul J. Larsen 
508 S. Tulane St. 
Albuquerque, N.M. 

William H. Rivers 
342 Madison Ave. 
New York 17, N.Y. 

Edward S. Seeley 
161 Sixth Ave. 
New York 13, N.Y. 

R. T. Van Niman 
4501 Washington St. 
Chicago 24, 111. 

Basic Research 
For Motion Pictures 



IT HAS BEEN POINTED OUT many times that engineering progress 
depends directly on basic research. Engineering is largely the 
application of scientific principles to the solution of practical problems. 
When new basic principles are discovered, engineering skill rapidly 
transforms these new ideas into things of value to both industry and 
consumer in the form of better or cheaper products. But when the 
available fund of basic information has been worked over for twenty 
years or more, the engineering "advances" are very apt to be reduced 
to gadgets, fancy knobs, and other trimmings which are of little real 
value even from a sales standpoint. 

The motion picture industry has for the most part left basic re- 
search to the various suppliers of film and equipment. Some re- 
markable improvements have come out of these industrial research 
laboratories, but it is only natural that their research should be aimed 
at results which will assist the commercial operations of the particular 
company operating such a laboratory. 

There are some subjects of vital interest to the motion picture in- 
dustry which are not directly related to any commercial product and 
therefore have not been the subject of research to the extent em- 
ployed in connection with directly salable products. Some such sub- 
jects are various physio-psychological effects on motion picture the- 
ater patrons. The effect of physical factors of the projected picture 
on eyestrain and fatigue is one, and the effects of various forms of 
illumination around the screen on the psychological reaction to the 
picture is another. 

It is obvious that basic research is slow and expensive, and cannot 
be expected to show immediate returns in cash or any other tangible 
manner. Certainly the SMPTE could not set up a research labora- 
tory for the study of such problems, although it is interested in such 
research and would like to encourage it in any way possible. 



Some of the older members no doubt recall that in the early 30's 
the Society did sponsor basic study on a fundamental problem. A 
Fellowship was awarded to Mr. Peter A. Snell for graduate study at 
the Institute of Applied Optics, University of Rochester. Mr. Snell 
studied visual fatigue in viewing motion pictures, an account of this 
study being published in the May, 1933, issue of the JOURNAL. While 
Mr. Snell made an excellent contribution to knowledge on this sub- 
ject, the possibilities along this line were not exhausted, and that was 
over sixteen years ago. 

Here seems to be a way in which the SMPTE can be of definite 
service to the industry. There must be numerous candidates for 
advanced degrees in physics at various universities, some of whom 
might be glad to devote their research to a subject relating to the 
physical and psychological aspects of vision. A list of proposed re- 
search projects prepared by the SMPTE might be instrumental in 
influencing a graduate student to undertake such work. The estab- 
lishment of some form of fellowship or honorarium for the research 
worker would no doubt also help. 

The views of Society members on this subject will be welcomed. 

Noise Considerations in 
Sound- Recording 
Transmission Systems 



Summary Noise limitations in sound-recording media are well known. 
With improved media such as magnetic materials, noise limitations imposed 
by the recording transmission system require consideration. Noise may be 
internally generated in the system or may be introduced from extraneous 
sources by electromagnetic coupling, or circuit exposures to interfering fields. 
Radio and audio frequency disturbances, crosstalk, thermal noise, shot-effect, 
microphonics, and alternating current hum are some of the interferences con- 

THE INHERENT noise limitations of various forms of recording 
media have been the subject of considerable study. The most 
commonly used materials in the past have been various types of disk 
records and photographic film. Unmodulated background noise in 
these materials is such that the available dynamic volume range be- 
tween noise and a maximum signal whose distortion is on the order 
of 2% is 30 to 40 db. This basic volume range has of course been 
effectively increased by use of such recording methods as: pre- and 
post-equalization, volume compression and expansion, dynamic noise 
suppressors which vary the bandwidth of the reproduced signal, and 
noise reduction devices as applied to photographic recording. The 
introduction of magnetically coated film has provided a medium 
capable of accommodating a volume range of some 60 db without 
recourse to complex electronic devices. With such an improvement 
in signal-to-noise ratio, noise limitations imposed by the recording 
transmission system require consideration in order that they may not 
become a limiting factor in over-all system performance. Noise may 
l)e internally generated in the transmission system, or may be intro- 
duced from extraneous sources by electromagnetic coupling, or may 
enter the system conductively, through circuit exposures to interfering 
fields. This paper will consider various types of externally and in- 

PRESENTED: October 14, 1949, at the SMPE Convention in Hollywood. 


130 F. L. HOPPER February 

ternally generated noise disturbances, which may seriously reduce 
the effective volume range of the recording media. It is of course 
recognized that there is an ultimate minimum of noise which cannot 
be reduced by any practical means, but it is the engineer's desire to 
approximate this limit. The various types of noise interference 
may be classified as outlined below. 


A. Radio Frequency Noise 

Disturbances due to radio frequencies entering the system are 
generally troublesome only when the signal is a modulated radio 
carrier. Subsequent demodulation in the system due to various 
factors may result in an audio frequency signal. Methods for correc- 
tion may require magnetic shielding or the use of circuit elements to 
provide a filtering action. 

This type of disturbance may also be introduced into the trans- 
mission system due to exposure of connecting circuits to interfering 
fields. Power connections to a-c mains, long microphone cables, or 
even the ground connection itself, may introduce such interference. 

Corrective measures depending upon the type of exposure may 
include the following: 

1. Use of magnetic shields 1 - 2 ' 3 ' 4 in low-level audio stages on trans- 
formers and vacuum tubes. 

2. Shielding of grid and grid return circuits in copper braid- 
shielded wire. 

3. Radio frequency filtering utilizing a small series choke coil of a 
few millihenrys inductance in the grid circuit as close as possible to 
the grid connection of the tube itself. A by-pass condenser of some 
100 micromicrofarads capacity connected from grid to ground may 
also be effective. Possible deterioration of the audio frequency re- 
sponse of the apparatus should be measured, when such filtering is 

4. Careful shielding of long connecting circuits such as micro- 
phone cables using a braided copper shield which is insulated from 
contact with the earth. 

5. Use of an electrostatically shielded transformer as a barrier 
between the primary power source and the system rectifiers, 


B. Audio Frequency Noise 

Audio frequency noises introduced by means of electric or magnetic 
induction are apt to consist principally of power line frequency or its 
multiples. In general, remedial measures are: increased physical 
separation between the source of disturbance and the affected 
apparatus; or the use of magnetic shields on either or both of the 
disturbing or affected equipment components. 

Audio frequency noise components introduced into the system via 
connecting external wires may also consist of power line frequency 
or its multiples. Use of transposed or twisted conductors is an 
effective means of reducing such interference. Such twisted con- 
ductors cause both sides of the circuit to be equally exposed to the 
interfering field, and the induced currents largely cancel. Magnetic 
shielding of such conductors is ineffective unless such magnetic 
materials as iron pipe or flexible conduit are used. Most effective, 
but seldom used, is a permalloy wrap similar to that used for continu- 
ous loading of submarine cable. 

Crosstalk may prove to be an undesirable source of interference 
and may be caused by: an inductive field, a shunt admittance or 
series impedance, or by a common impedance. 

In addition to shielding, crosstalk may be reduced by using trans- 
posed or twisted transmission pairs. The twist, to have maximum 
effect, should be completely uniform, a condition which generally 
obtains when each wire of the pair is of equal length. If the inter- 
fering induction field affects all parts of the circuit with equal inten- 
sity only a loose twist is required. If all parts are not affected with 
equal intensity, a tight twist is indicated. For example, if the cross- 
talk occurs between adjacent pairs in a cable or form, since the cable 
is probably short in comparison with the wavelength of the interfering 
induction noise, only one transposition is theoretically required since 
each half of the circuit is equally exposed. If, however, the circuit 
being interfered with passes near a point source of interference, the 
intensity of the field will diminish as the square of the distance from 
the point. In such a case there should be an infinite number of 
twists per unit length to annul completely the interfering field. The 
principal consideration is that both sides of the circuit must have 
exactly equal exposure to the interfering field. It must be realized 
that the induced currents on each wire of the pair, although they may 
be exactly equal in amplitude and opposite in phase, cannot cancel 

132 F. L. HOPPER February 

until the currents flow into a common junction. At this point, which 
should be grounded, complete cancellation results only if the currents 
are exactly equal and in phase opposition. 

Both balanced and unbalanced types of transmission systems are 
in common use. Balanced-to-ground transmission has been found 
by far the most practical when circuits are long and is useful when the 
circuits are of the order of 20 ft in length. However, expensive coils 
and more complicated components are required. Unbalanced cir- 
cuits require less complicated components but because of the un- 
avoidable unequal series impedance they are more susceptible to 
inductive effects. The best results will be obtained from either 
type of transmission when proper attention is given to grounding. 
The choice of type of transmission system can be made only after 
carefully weighing the economies of the unbalanced method against 
the higher protection factor of the balanced method. 

Level differences in cables and wiring forms should be kept to 
within a reasonable figure. A well-shielded cable form can tolerate 
level differences exceeding 60 db between conductors if the form is not 
longer than 10 ft. This requires intelligent shielding and circuit 
geometry. Under these conditions the crosstalk may be 60 db or 
more down from average signal levels when unbalanced transmission 
is used and considerably more when balanced transmission is used. 
Differences in level in cable forms of 60 db or greater should be tol- 
erated only if it is a local condition as in the case of equipment within 
a rack or console. Interconnecting circuits between racks should 
be kept physically separated, in separate conduits if possible, when 
level differences are 60 db or greater, or when both balanced and un- 
balanced circuits are in the same cable forms. The individual pairs 
of these cable forms should have separate copper braid shields and 
the braided shields themselves should be insulated from each other. 
When crosstalk occurs in a cable form despite all precautions it is 
often found that by selecting pairs the effect can be reduced to a 
tolerable level. This also applies to lead cable. If the crosstalk 
occurs between circuits, one or both of which is unbalanced, the only 
effective remedy, outside of increased physical separation, may con- 
sist in balancing the circuit by means of a repeating coil. If unbal- 
anced transmission must be restored as the circuit pair leaves the 
form, two coils may be required. If the crosstalk is not severe, an 
elaborately balanced coil will not be required. When exploring for 
crosstalk a relatively high frequency should be used, since crosstalk 



more apparent at these frequencies. Circuit irregularities are 
sometimes such that a narrow region of crosstalk occurs and perhaps 
a thousand cycles or so on each side of this region the crosstalk is 
not detectable. It is usually good practice to short-circuit or ground 
all unused pairs in a cable form. 


.1 . Thermal Agitation Noise 

The current in the output of an amplifier due to the thermal 
agitation noise in a resistance R connected to the input terminals of 
the amplifier is 5 : 

/. = T 


where I = current in output of amplifier 

R(u) = resistance connected to input of amplifier 
F(a>) = transfer admittance of amplifier = I/V 

V = voltage required across amplifier input terminals to cause a current = 
/ to flow in the output '. 

CO = 27T/ 

T = degrees Kelvin 

A' = Boltzmann's gas constant = 1.37 X 10 ~ 16 joules per degree. 

If R is independent of frequency, usually a valid assumption over the 
audio frequency range, and if the amplifier is flat over the audio band 
from /i to / 2 and has zero gain outside of these limits, the above equa- 
tion becomes : 

F 2 = 4KTR (f t - / t ). 

This voltage V, is that voltage of any frequency between / 2 and f\ 
which, if applied to the input of the amplifier, would cause a current 
1 to flow in the output of the same magnitude as does the thermal 
agitation noise in R. Putting T = 300 K (80.6 F), letting (/ 2 - /i) 
equal 15,000 cycles, the equation becomes: 

F 2 = 2.46 X W~ U R. 

The voltage squared in the above equation is, however, the open 
circuit voltage across the resistance R. It can be shown that the 
conditions under which the amplifier operates are such that the resist- 
ance R is actually terminated in its own impedance which reduces 

134 F. L. HOPPER February 

the thermal agitation voltage by a factor of 2 2 or 4. F 2 then be- 

F2 = 2.46 x 10 ~ 16 R 


V 2 = .615 X 10~ 16 R 
and V 2 /R = .615 X 10 ~ 16 watts. 

From this it can be seen that the thermal agitation noise is a constant 
power of 132 dbm (under the conditions prescribed above) and 
that the voltage consequent to this power is determined by the 
resistance alone. The amplifier should be visualized as an ideal 
transformer terminated on its primary with the generator resistance 
R, and on its secondary by the load resistance of the external circuit. 
The load resistance divided by the impedance ratio of the ideal trans- 
former presents a termination to the generator resistance equal to it- 
self. The open circuit thermal agitation voltage is consequently 
reduced to one half. 

If the internal impedance of the amplifier is very high with respect 
to the generator impedance a justified assumption is that all of the 
thermal agitation noise in the output comes from the generator. The 
noise in the output, provided the amplifier is designed so that there 
is no contribution of noise from the input tube, will then be 132 
dbm plus the gain of the amplifier. 

The noise due to thermal agitation is irreducible by any practical 
expedient. Any noise originating at points subsequent to the grid 
of the first tube represents an addition to this noise without a corre- 
sponding increase in signal. The limit to be attained 6 in signal-to- 
noise ratio is therefore the ratio of the input signal to the thermal 
noise of the input circuit. 

B. Shot-Effect and Other Tube Noise 

Shot-effect 7 is due to the fact that the electron flow to the anode of 
a vacuum tube is not of uniform density. Space charge in vacuum 
tubes creates a cushioning effect which reduces the shot noise by 
increasing the uniformity of electron flow. Theoretically, in the 
case of complete temperature saturation, the space charge would 
reduce shot-effect to zero. Practically, this condition obtains in 
vacuum tubes used in most amplifiers. Formulas exist 8 which allow 
accurate calculation of shot-effect both in the presence and absence of 
space charge. 


The major noise contribution l>y the vacuum tul>< is the, infernal 
thermal agitation noise. 9 This occurs in the plate circuit. In am- 
plifier design work it is important to evaluate this noise so that the 
first amplifier stage will sufficiently amplify the input thermal noise 
to override the thermal noise of the tube itself. If -this course is 
followed the tube contributes little noise to the circuit. The thermal 
noise arising in the tube comes from the cathode-to-plate resistance 
within the tube itself. It is the same in all respects as thermal noise 
generated in a resistor but it is important to remember that the 
temperature of the plate resistance is that of the cathode. In modern 
tubes this temperature is of the order of 1000 to 2000 K. Because 
the plate resistance appears in shunt with the external plate load 
resistance, the total effect must be considered to be that of two in- 
dependent generators in parallel. Formulas for calculating this noise 
are fully explained elsewhere. 8 

There are other sources of noise within a vacuum tube due to 
flicker effect, collision ionization, secondary emission, positive ion 
noise, and noise which may be generalized under the heading of micro- 
phonics. Except for the latter, which is a primary noise factor under 
certain conditions, the noise contribution by such secondary effects 
is negligible. 

The gong-like sound of microphonic noise 10 is familiar. Tubes 
may be designed to minimize these effects, but in any case may re- 
quire selection. Other means of reducing microphonics include shock 
mountings where the period of the tube and socket assembly is in- 
creased to reduce susceptibility to external mechanical shock; or if 
the shock is airborne, surrounding the tube with an acoustic shield. 
Another type of noise which is classed under the heading of micro- 
phonic noise is termed "sputter noise." This noise arises due to 
mechanical support problems of the tube elements and internal in- 
sulation leaks. Some types of sputter noise, such as those affected 
by element support considerations, are only apparent in conjunction 
with the usual microphonic noise and will be eliminated or reduced by 
the same methods which reduce microphonics. The other type, due 
to insulation leakage, may appear independently and since this is a 
tube manufacturing problem the engineer can only select tubes free 
from this noise. 

Noise originating in vacuum tubes is often considered, from an 
amplifier design standpoint, to be equivalent to a thermal noise gen- 
erated in a fictitious resistance 8 in the input. While tube noise has a 

136 F. L. HOPPER February 

frequency characteristic, and thermal noise has none, the convenience 
of the assumption outweighs the defect in the analogy. This ficti- 
tious resistance called the "equivalent noise input resistance" is meas- 
ured as follows. 11 Assume that a resistance equivalent to the internal 
resistance of a generator, such as a microphone, is connected through 
a suitable transformer to the grid of a vacuum tube. The thermal 
noise arising in the resistance will be amplified by the tube in the 
normal way. This noise represents the lowest possible noise in a 
system of given frequency response. The tube also generates noise 
within itself which adds to the external thermal noise. Maximum 
signal-to-noise ratio can be achieved only when tube noise is negligible 
as compared to thermal noise. If the grid of the tube is grounded 
and the external equivalent generator resistance and coupling trans- 
former are disconnected, the noise currents remaining in the output 
of the tube will be due to the tube itself. This noise is measured by 
a voltmeter or thermocouple having square law characteristics. 
Additional amplification in conjunction with the meter will be re- 
quired. If the short circuit from grid to ground is removed and a 
resistor placed across the same points, a value of resistance will be 
found which, due to its thermal noise output, will exactly double 
the meter deflection (or raise it 3 db) . This is an indication that 
the thermal noise generated by the resistor is equal to the noise 
generated by the tube. The value of the resistance is called the 
"equivalent noise input resistance" (/2 e m) of the tube. The R eni of 
tubes can be calculated and formulas are available for triodes, pen- 
todes, triode and pentode mixers and multigrid converters. 

Tubes showing extremely low values of R eni do so by virtue of a 
high transconductance usually obtained by close grid-to-cathode 
spacing. Consequently there are other considerations, such as 
microphonic characteristics, stability of performance, and low hum 
disturbance factor, in selecting a tube for audio pre-amplifier use. 

C. Fluctuation Noise in Granular Type Resistance Carrying Direct 

When direct current is passed through carbon resistors which are 
granular in nature or through resistors consisting of sputtered or 
evaporated metal films, small voltage fluctuations appear across the 
resistor terminals. This adds to the basic noise of the system and any 
attempt to reduce the basic noise to as low a value as possible must 
take this into consideration. This has been termed contact noise and 


the hypothesis advanced ' " is that it is due to minute i< 
variations in the region of contact het ween the granular boundaries. 
As would be expected the expression for the generated voltage is 
complicated. The vulture depends upon the value of the resistance, 
the direct-current voltage across the resistance, the frequency range, 
and two empirically determined exponents and a constant. The 
dist ribution of noise with frequency shows an increase in noise energy 
with decreasing frequency at the rate of 3 db per octave. 

D. Hum Noise Due to A-C Operation of Filaments 

The use of indirectly heated unipotential cathodes removes most 
of the factors which produce hum 13 in vacuum tubes when operated 
from an alternating-current source. Due to the almost universal 
use of heater-type cathodes no consideration of filamentary cathodes 
will be discussed. Hum is due to (1) the magnetic field produced 
by the heater current; (2) leakage resistance from heater to other 
electrodes; (3) electrostatic fields due to lack of complete heater 
shielding; and (4) electron emission from the heater. 

The magnetic field creates a double-frequency hum. The problem 
is one which must be dealt with in tube manufacture. Remedial 
measures consist in "twisting" the heater wires or other such methods 
of obtaining essentially noninductive heaters, high-voltage-low-cur- 
rent heater operation and certain modifications of the tube geometry. 

Leakage resistance from heater to other electrodes is often due to 
"getter" deposits within the tube. This type of hum can be reduced 
by lowering the impedance of the affected electrode to ground. 
Another method, when leakage exists between heater and cathode, 
is to bias the heater either positive or negative with respect to the 
cathode. A =>=10-volt bias is almost always adequate and little 
difference is usually found between a positive or negative bias. 

Hum due to capacitive coupling is of fundamental power main fre- 
quency. The most objectionable capacity is that which exists 
between heater and control grid. The effect of this capacity is 
reduced by lowering the grid-to-ground impedance which is another 
reason that the secondary impedance of an input transformer should 
be kept as low as possible. A "bucking-out" potentiometer is some- 
times used. This consists of a potentiometer of several hundred 
ohms connected across the heater leads, the potentiometer swinger 
being connected to the grid-return lead. 

138 F. L. HOPPER February 

Hum due to heater emission can occur when any electrode is at 
higher potential than the heater. The situation is sometimes ag- 
gravated by some of the cathode coating (barium and strontium 
oxides for example) getting on the heater during tube manufacture. 
The work function of the emitting surface of the heater for emission 
currents can thereby be greatly reduced. The solution most em- 
ployed is to bias the heater positive with respect to the affected ele- 
ment, usually the cathode. Ordinarily a volt or two is adequate. 
In rare instances the emission may be from cathode to heater in which 
case a negative bias is required. 

Referring specifically to magnetic recording, since it is necessary 
in reproduction to employ equalization at low frequencies at a rate 
of 6 db per octave, in order to achieve uniform over-all frequency 
response, low frequency noise disturbances assume considerably 
greater importance. For example, if some 25 db of equalization at 
50 to 60 cycles is required for flat response, any fundamental power 
line frequency disturbance must be reduced by such an amount, if 
signal-to-noise ratio is to be maintained. 

Noise interferences causing the greatest difficulty are: 

(1) Introduction of power line fundamental or multiples thereof, 
by magnetic coupling to apparatus components; (2) introduction of 
power line frequency or multiples by power wiring to exposed com- 
ponent or interconnecting wiring; and (3) internally generated power 
line multiples due to a-c operation of vacuum tube heaters. 

Corrective measures as indicated in previous sections of this paper 
will include : use of magnetic shields on both the interfering appara- 
tus element and the equipment component receiving the interference; 
use of shielded and twisted wire pairs both in internal and intercon- 
necting wiring; and possible recourse to rectified alternating current 
or use of direct current on the heater elements of vacuum tubes in 
low-level amplifier stages. 

Other described forms of interference are probably no more severe 
with magnetic systems than those using a photographic medium. 

In systems designed with the described noise factors in mind, it 
will generally be possible to fully achieve the signal-to-noise possibil- 
ities inherent in the chosen recording medium. 



(1) A. G. Ganz and A. C. Laird, "Improvements in communication trans- 
formers," Bell Sys. Tech. J. t vol. 15, p. 136; January, 1936. 

(2) W. G. Gustafson, "Magnetic shielding of transformers at audio frequen- 
cies," vol. 17, p. 416; July, 1938. 

(3) Samuel Levy, "Electromagnetic shielding effect of an infinite plane con- 
ducting sheet placed between circular coaxial coils," Proc. I.R.E., vol. 24, p. 923; 
June, 1936. 

(4) Walter Lyon, "Experiments on electromagnetic shielding between one and 
thirty kilocycles," Proc. I.R.E., vol. 21, p. 574; April, 1933. 

(5) J. B. Johnson, "Thermal agitation of electricity in conductors," Phys. 
Rev., vol. 32, pp. 97-109; July, 1928. 

(6) J. B. Johnson and F. B. Llewellyn, "Limits to amplification," Bell Sys. 
Tech. J., vol. 14, p. 85; January, 1935. 

(7) "The theory of the Schroteffekt," J. Frank. Inst., vol. 199, p. 203; Febru- 
ary, 1925. 

(8) F. B. Llewellyn, "A study of noise in vacuum tubes and attached cir- 
cuits," Proc. I.R.E., vol. 18, pp. 243-265; February, 1930. 

(9) G. L. Pearson, "Fluctuation noise in vacuum tubes," Bell Sys. Tech. J., 
vol. 13, p. 634; October, 1934. 

(10) D. B. Penick, "The measurement and reduction of microphonic noise in 
vacuum tubes," Bell Sys. Tech. J., vol. 13, p. 614; October, 1934. 

(11) J. J. De Buske, "Noise measurements in vacuum tubes," Bell Labs Rec., 
p. 456; August, 1943. 

(12) C. J. Christensen and G. L. Pearson, "Spontaneous resistance fluctuation 
in carbon microphones and other granular resistances," Bell Sys. Tech. J., vol. 15, 
p. 197; April, 1936. 

(13) J. O. McNally, "Analysis and reduction of output disturbances resulting 
from the alternating-current operation of the heater of indirectly heated cathode 
triodes," Proc. I.R.E., vol. 20, p. 1263; August, 1932. 

Photography in the 
Rocket-Test Program 



Summary- The vast bulk of information gained during a rocket or 
guided-missile test firing at Inyokern is obtained photographically. This 
type of detailed recording requires a large number of specialized photographic 
instruments of various types. They include cinetheodolites, ribbon-frame 
cameras, high-speed motion picture cameras, and others. The importance 
of color in this program is described, together with some of the difficult prob- 
lems that are encountered when combining color film with the high shutter 
speeds necessary to stop the motion of fast-moving test objects. Some solu- 
tions to these problems of underexposure are mentioned. 

THE PRIMARY FUNCTION of the United States Naval Ordnance Test 
Station is the development and testing of weapons, particularly 
those in the fields of rockets and guided missiles. Located west of 
Death Valley in California's Mohave Desert, the 800-square-mile 
expanse of desert and mountains that contains the principal test 
ranges was selected not only for its isolation but also for its excellent 
seeing conditions, for the vast bulk of information obtained from these 
test firings is photographic in nature. 

While many of our photographic problems remain unsolved, we 
feel that we have come a long way since those hectic wartime days 
when scientists and technicians from the California Institute of 
Technology first pitched camp on the desert sands and commenced 
firing rockets from crude launchers. Those pioneers came armed 
with only a few well-worn Eyemo and Bell and Howell Superspeed 
cameras as their means of securing information on the test firings, 
but they developed many of the basic principles of ballistics photog- 
raphy as it is practiced today on our numerous, well-instrumented 
testing ranges. That early type of photographic coverage served 
only as a more permanent record of the same information obtained 
visually by the observers; now an amazing variety of specialized 
instruments furnish precise numbers denoting velocity, acceleration, 

PRESENTED: October 10, 1949, at the SMPE Convention in Hollywood. 


range, altitude, spin rate, and other data that are the basis of any 
scientific evaluation and improvement program. 

One of the most valuable of these specialized instruments is the 
cinetheodolite, recording missile position against time in a network 
that permits the computation of accurate trajectories of flight. At 
Inyokern, the basic cinetheodolite is the German Askania Kth 41, as 
modified. This instrument operates at 2 or 4 frames per second, 
producing double-frame 35-mm pictures on Ansco Color motion pic- 
ture film. Color film has been found to be the most suitable means of 

Fig. 1 . Askania Cinetheodolite used for guided missile instrumentation 
at Naval Ordnance Test Station. These instruments have been modified 
as shown to permit one-man tracking of high-speed missiles. 

obtaining easily assessable contrast between painted rockets and the 
blue-sky background. The Askania theodolite has been improved at 
Inyokern to permit free-wheeling single-operator tracking in place of 
two-man geared tracking to permit fast-moving test objects to be 
followed. Since high-tracking rates resulted in blurred images of the 
azimuth and elevation dials, the shutters exposing these dials have 
been removed and replaced by gaseous-discharge lamps, which effec- 
tively "freeze" the dial motion. Work has also been done to achieve 
more effective synchronism of the several instruments in use than was 
possible with the original equipment. 




The Bowen Ribbon-Frame Camera, described in detail by Green and 
Obst,* is a precision nontracking camera capable of providing ex- 
tremely accurate metric data on 5V2-inch aerial roll film at rates of 
from 30 to 180 frames per second. It is a valuable supplement to the 
Askania theodolite, especially during the early portions of flights 
through burning time. 

Fig. 2. Typical sample of record obtained by Askania Cinetheodolites. At 
the top of the frame are, from left to right, azimuth scale, frame number, and ele- 
vation scale. 

Thirty-five-millimeter Mitchell cameras are operated at speeds up 
to 120 frames per second to provide information on roll, pitch, yaw, 
separation of booster from missile, and off-range deflection. The 
addition of a second lens and shutter into the side of the camera case 
has converted these cameras into so-called "chronographs." This 
auxiliary camera system photographs the image of a stop watch on 
the corner of each frame of film, permitting the assessment of the 
action photographed to the nearest Yioo second. Forty-inch lenses 
are most commonly used with these cameras, which are operated 

* See pp. 515-23, November, 1949, Jour. SMPE. 


PHOTOGRAPHY IN ll<>< KI .1 Ti i 


from conventional tripods or from powered I nick in ; mounts. As in 
the case of the Askania theodolites, Ansco Color film is used in this 
camera on most firings. 

To obtain detailed slow-motion studies of Immrliings, separations, 
static firings, or detonations, extensive use is made of 16- and 35-mm 
high-speed motion picture cameras operating at speeds up to 4000 

Fig. 3. The Bowen Ribbon Frame Camera can be accurately aligned 
to photograph an expected rocket or missile trajectory. The camera 
operates at from 30 to 180 frames per second, recording on 5V:rin. aerial 
roll film. 

frames per second. Some success has been achieved at Inyokern 
using color film at speeds up to 500 frames per second, because of the 
brightness level usually encountered on the test ranges which are 
located on alkaline dry-lake beds. 

These and other optical methods of obtaining direct photographic 
data of test firings are no more important, however, than the indirect 
photographic records obtained electronically. Recording cameras 




photograph oscilloscope tubes, record magnetic pickup traces, and, as 
oscillographs, receive vital telemetering data from the test objects 
themselves. Their operation adds to the thousands of feet of motion 
picture film, wide film, and wide paper which is expended to obtain 
detailed recordings of a single flight that might last only a few seconds. 

Fig. 4. Typical sample of record obtained by Bowen 
Ribbon Frame Cameras. This record shows an 11.75- 
in. aircraft rocket striking and penetrating an armor 
plate target with resultant high order detonation. 

Less quantitative but no less valuable information is secured by 
"newsreel" photographers who obtain 16-mm Kodachrome records of 
the assembly and checkout of missiles, their placement on the 
launchers, their flights, and the state of their remains, when found. 
This coverage is supplemented by remotely controlled 4- X 5-inch 
still cameras which obtain Ektachrome photographs of missiles as 
they just clear the launchers during firing. 




Fig. 5. Mitchell Chronograph 35-mm Tracking Camera with 18-in. 
lens used at Naval Ordnance Test Station to determine missile attitude, 
roll, and separation information. 

Fig. 6. Typical sample of film obtained by 35-mm Mitchell Chronograph camera. 
Stop watch image is recorded in corner of each frame. 




The Naval Ordnance Test Station conducts extensive work in the 
field of airborne rocket firings and underwater ordnance studies which 
supplement the firings performed on the ground ballistics ranges. At 
Inyokern many interesting adaptations of camera equipment have been 
made to obtain data on firings made from planes in flight against 
ground targets, while the Underwater Ordnance technicians in Pasa- 
dena are busy developing totally new techniques of high-speed under- 
water illumination and photography. 

Fig. 7. This MK 45 powered tracking mount carries a 16-mm Eastman 
High Speed Camera with 20-in. lens and a 35-mm Mitchell Chronograph 
Camera with 40-in. lens. 

The photographic operations on the ranges and in the research 
laboratories funnel into the Photographic Laboratory Section, which 
operates an 8000-square-f oot production laboratory in one wing of the 
huge new Michelson Laboratory. While the amount of footage in- 
volved is not large by commercial laboratory standards, the lengths of 
film handled are just as valuable as those shot on a big production 
scene in an epic Hollywood film. This film is strictly a one-take 
affair there are no retakes. An atmosphere of urgency is ever pres- 
ent in the processing laboratories; frequently the test conductor is 
standing by to see the first round's records before proceeding with the 
day's firing schedule. Yet the film must be handled with the extreme 




care required to produce records that can be minutely assessed under 
powerful magnification to deliver metric data. 

The use of color film in the 'rocket instrumentation program pre- 
sents several difficulties. Long focal-length lenses are usually required 
to photograph distant objects in space or to photograph objects on 
the ground from safe distances. Such lenses are usually rather slow, 
and high shutter speeds must be used to freeze the motion of super- 
sonic missiles. This combination of factors usually means severe 

Fig. 8. Two 16-mm Eastman High Speed Cameras are aligned to cover the 
launching of a guided missile. The radio receiver at left receives a broadcast 
1000-cycle note and transmits it to timing lamps in the cameras. 

underexposure when a film with an American Standards Association 
rating of 10 is used. Therefore, the Photographic Laboratory has 
devoted considerable study to methods of effectively raising the 
emulsion speed of material such as Ansco Color Type 235 motion 
picture film and Ektachrome cut film. Efforts to date have been 
confined to a juggling of recommended developing times, although it is 
realized that several other methods are available for investigation. 
At present, we obtain normal Ektachrome transparencies exposed at 
//2.5, Vsoo second, by increasing the first developer time from 15 to 


20 minutes. Much of the Mitchell Chronograph Ansco Color motion 
picture film is exposed at //8.0, l / m second. We have increased 
Ansco Color first developer time to as much as 33 min and have 
also juggled color development times. These deviations in process- 
ing times do not appear to have caused marked changes in color 
balance, although color fidelity is not of great concern in our case. 
Rocket photography at Inyokern has indeed come a long way since 
those pioneering days of the California Institute of Technology. Yet, 
we feel that we are only beginning to learn the science of ballistics pho- 
tography. As rmssiles become faster and go farther, instrumentation 
difficulties multiply, and it becomes evident that much remains 
unknown and untried. As workers in the program of our nation's 
defense, we appreciate the valuable assistance and advice given us by 
the Society of Motion Picture Engineers as we attempt to add to the 
sum of knowledge in our still infant field of specialized photographic 

High-Speed Processing 
of 35-Mm Pictures 



Summary A piece of apparatus is described by means of which it is possi- 
ble to develop, fix, wash, and dry pictures on 35-mm film in four seconds. 
Integral with the apparatus is a camera lens with which the pictures are 
formed and a projection system with which they are projected. Though the 
particular device has been designed for radar photography, it is believed to 
have a variety of other uses and applications. 

THE TITLE of this paper may be misleading to the motion picture 
engineer for whom the concept of speed of processing is based 
upon the idea of feet of film per minute. The system we are about to 
describe in its simplest form is capable of only a small footage 
output per unit of time. This output of approximately one foot of 
35-mm film per minute would be regarded as extremely slow by those 
engaged in the quantity production of release prints. 

The system and equipment are, in another sense, capable of the 
highest speed of photographic processing with which the authors are 
acquainted. In this narrower sense, we refer to the time lapse occur- 
ring between the instant when a silver halide-gelatin latent image is 
subjected to the necessary treatment, and the final production of a 
usable silver image. Regarding this silver image, we impose some 
arbitrary qualitative specifications: that it shall be free from undis- 
solved silver halide, meaning that it shall be completely "fixed"; 
that it shall be permanent, meaning that the products of 'chemical 
action tending to produce image deterioration shall be absent; that 
it shall be dry, and that its over-all quality, as defined by density, 
contrast, inertia speed, uniformity, and maximum density, shall be 
favorably comparable to results achievable by conventional process- 

Neither the demand for developments in this field of maximum, 
rapid processing nor the efforts in cutting down processing time have 
been lacking. Portraits "while you wait" have been available at 
county fairs since 1910. Better portraits while you wait a shorter 

A CONTRIBUTION: Submitted November 16, 1949. 


150 TUTTLE AND BROWN February 

time are sold by the Photomaton. The Land-Polaroid system re- 
quires the shortest wait and gives the best of results in the hands of the 
amateur. These developments have reduced processing and drying 
time to the order of one minute. 

During the latter part of the War, the demand for quick processing 
of radar image pictures was met by the Eastman Kodak Co. which pro- 
duced a unit that photographed the radar screen quickly, processed 
the image, and projected it onto a screen. The image was made on 

Fig. 1. Photograph showing film path through the apparatus. 

16-mm film and occupied an area equal to the largest circle that could 
be inscribed in a single 16-mm frame. Only a small portion of the 
total film area was used. The image appeared on every fifth frame 
and the equipment was capable of producing complete pictures at 
13.5-sec intervals. A projected picture, seven feet in diameter, of 
the Plan Position Indicator Radar Scope was the end result. 

The cycle time of this unit made it possible to keep up with 4 rpm 
radar. The picture size, definition, and brightness were adequate to 


resolve and present all of the data shown on the radar set which the 
photographic unit was designed to complement. 

Advances in radar techniques have resulted in higher scanning 
rates and higher orders of data resolution. The requirements, there- 
fore, for the associated photographic equipment have, as a result, 
been made more difficult. Scanning rates of 15 rpm demand a proc- 
essing time of four seconds. Resolving power and other considera- 
tions require a 10-ft diameter picture of higher brightness than that 
produced formerly. 

The apparatus we are about to describe has been developed to meet 
the new requirements and also to meet them in push-button control 
manner such that one entirely unacquainted with photographic 
techniques can operate the device. 


Thirty-five-mm perforated film is used in the apparatus. Figure 
1 is a plan view of the film path. This film is fed from a 400-ft 
retort, past a picture taking station (1), a processing station (2), a 
projection station (3), over an intermittent sprocket, and onto a take- 
up reel. The block labeled B in Fig. 1 performs a multiplicity of 
functions and since it is the heart of the system it is deserving of some 
detailed description. 

Figure 2 is a section of this block, drawn out of scale to illustrate 
some of the unique principles of this quick photography method. 

First position: The film (1), shown edgewise, passes first the latent 
image forming position (2), where the optical image, 0.700 in. in 
diameter, is brought to focus on the emulsion through the transparent 
base. At the end of the exposure period, the film is advanced the 
distance of five standard perforations (0.935 in.) to the processing 
position (3). 

Second position: Within 2 milliseconds of its advance, a vacuum of 
about 4 in. of mercury is applied through an annular ring (4). This 
vacuum causes the film to be sucked inward (downward in the figure), 
resulting in the formation of a paraboloidal surface convex toward the 
emulsion side of the film. Matching this emulsion surface but re- 
moved from it by 0.005 in. is a polished stainless steel surface (5). 
The two surfaces emulsion and steel form the boundaries of a 
parabolic shell. The vacuum is formed and the film is sucked inward 
only when a poppet valve (6) is opened. 

152 TUTTLE AND BROWN February 

Now, when a second valve (7) is opened, a solution piped through 
the tube (8) is sucked through the central orifice (9). Under the in- 
fluence of the vacuum, a chemical solution passes with high and 
nearly constant velocity of 18 cms/sec, radially across the film surface 
and out the annular aperture (4). 

Important to the achievement of fast and uniform development, 
fixation, washing, and drying, is the accurate control and continual 
replenishment of fluids as they flow in contact with the emulsion 
surface. The development of the shape of the paraboloidal shell 
formed between film and the stainless steel faces to give the opti- 
mum conditions, which though an interesting problem in hydraulics and 
chemistry, is considered outside of the scope of the present discussion. 

What has been described in the foregoing is the action of two valves, 
one to produce a vacuum, and the second to admit a solution. In the 
present equipment three other valves are used, one below and two 
above the plane shown in Fig. 2. 

Each of these valves opens to a plumbing line similar to (8) in 
Fig. 2. Subsequently, in the device we are describing, valves open 
as follows : to developer, to fixer, to wash solution, and to air. The 
fluids mentioned are each in turn admitted through the center orifice 
and sucked out through the annular ring aperture. 

Thus in action, developer, fixing bath, wash solution and air chase 
each other out in rapid succession. For the most rapid chemical 
performance, high velocity of movements and minimum amounts of 
each solution present in the shell are prerequisite. In this equipment 
less than 0.01 ml of solution is sufficient to fill the shell and this solu- 
tion can be displaced or evacuated in 1 millisecond. 

Third position: Upon completion of processing and drying, the 
film moves again to the station (10) where it enters a projection opti- 
cal system. This system, which is an optical relay, has the final 
field lens (11) contained within the block. 

The film as it enters this position has had all of the superficial liquid 
from the last processing solution removed. There remains, however, 
a small amount of absorbed moisture in the emulsion layer. When 
any appreciable amount of radiant energy passes through the image 
at this point, the absorbed moisture gradually "sweats" out of the 
emulsion. Some means must be provided to disperse this moisture 
and a simple and adequate means is provided by allowing low-pressure 
air heated by the projection lamp to flow over the film surface. As 




indicated in the figure, this air enters the lens tube at (12) and passes 
around the lens (11) through a channel (13). 

This completes a brief description of the principal functions of the 
heart of the apparatus (block B). Another necessary function of the 
block is the immediate pre-heating of the film area to be processed 
and of the solutions immediately before they are used. 

The use of elevated temperatures to increase processing rates is 
well known and good use of this effect was made in the case of the pre- 
ceding fast processing equipment. In this connection, it should be 
pointed out that there is a definite maximum safe temperature for both 

Fig. 2. Diagram of processing device. 

processing solutions and film. If, because of the design of the proc- 
essing equipment, some cooling of the solutions is necessary during 
the course of processing, the average processing temperature must 
perforce be less than the maximum tolerable value. Under such con- 
ditions, the processing cycle needed to reach a prescribed photo- 
graphic result cannot be a minimum. To obtain this minimum time, 
it is obvious that temperatures should be kept at the safe maximum 
throughout the entire cycle. 

It is to accomplish this end that into the design of block (B) have 
been built all the necessary heating and control units to keep both 




film and solutions at the highest tolerable temperature throughout 
the processing cycle. The maze of plumbing, the valves, valve 
actuators and the thermostatically controlled heaters are, accordingly, 
combined in a single unit. Since the chemical activity of the solutions 
used must be at maximum, the materials with which these solutions 
come in contact must be inert, consequently the whole unit has been 
machined from a block of stainless steel (18-8 Mo). 

Fig. 3. Photograph of processing block with photographic station 
and projection lamp house. 

This unit (shown in Fig. 3), though it is small enough to hold in the 
hand and weighs less then 3 Ib, contains the following elements: 

A film track and cover plate, 

A 35-mm film processing machine, 

A 35-mm projection-condenser optical system, 

Solution heaters with thermostat control, 

Solution control valves, 

A film dryer. 


Since a prime purpose in the development of this equipment has 
been the reduction of processing time 'compared to that of the pre- 
viously described fast processing equipment, a tabular comparison of 
the results may be of interest. The original cycle compared to the 
present cycle is shown in Table I. 


(Patent 2,446,668; and present equipment Patent Applic. 114,701) 


time, sec 

time, sec 

Start developer flow 



Stop developer flow 



Start developer removal 


Finish developer removal 


Start fixer flow 



Stop fixer flow 



Start fixer removal 


Finish fixer removal 


Start wash flow 



Stop wash flow 



Start wash removal 


Stop wash removal 


Finish drying 
Start pull-down 



Finish pull-down 



It should be noted that, for the same extent of development, 4.3 
sec are required by previous equipment and 0.8 sec by present equip- 

For complete fixation 5.4 sec are required by the previous equip- 
ment compared to 1.7 sec for the present equipment. 

With previous equipment 0.9 sec are allotted to washing; with 
present equipment 0.3 sec are allotted to washing. 

In the previous equipment 0.6 sec are allotted to drying. In the 
present equipment 1.18 sec are allotted to drying. This results in a 
dryer film for projection with the result that the picture in the new 
equipment shows much less tendency to weave and warp. 

In the previous equipment, 0.5 sec are required for pull-down. In 
the present equipment this time is reduced to 0.01 sec. The principal 
advantage here is concerned with the fact that during film move- 
ment data presented by the radar is lost. This loss with the pres- 
ent equipment is only one-fiftieth the loss with the previous equip- 




The more efficient chemical results achieved with the present equip- 
ment are due principally to two factors: (1) greater agitation of the 
solutions resulting from faster liquid movement, and (2) higher aver- 
age solution temperature that can be maintained during the process. 

The equipment uses 0.75 ml of developer, 1.2 ml of fixing solution 
and 0.5 ml of water per picture. 


Fig. 4. General view of the camera-processor-projection unit. 

Since there is nothing particularly new in the equipment aside from 
the parts of the system just described, description of the rest of the 
unit will be brief. 




The whole unit, of which Fig. 4 is a photograph, consists of four 
sectional housings. 

A is a mounting for an 8X12 in. front-surface mirror whose func- 
tion it is to turn the projected beam horizontally. The inclination 
of this mirror is adjustable by fine thread screws. This adjustment 
is for the purpose of centering the screen image. 

B is a housing for the optics, light source, processing block, film 
supply, and take-up. 

C contains the solution supply tanks. 

D contains the vacuum pump and disposal tank. 

Fig. 5. Solution supply tank. 

Figure 5 shows three Lucite solution supply tanks, each of 1.5 gal 
capacity, and each equipped with electric solution level indicators. 
Solutions are piped from these tanks to the processing head with 
J/8 in. Saran tubing. 

158 TUTTLE AND BROWN February 

Figure 6 shows the interior of the lower unit with its Lucite waste 
tank of about 6-gal capacity. It also is equipped with a warning de- 
vice to indicate its state of fullness. 

Cycling Control System 

Figure 7 shows the cycling control system. The right-hand cam 
shaft carries the necessary cams and microswitches to control all 
processing functions. The solution valves are solenoid-operated and 

Fig. 6. Waste tank and vacuum system. 

the cams are timed as indicated in the right-hand column of Table I. 
The left-hand cam shaft, driven by a synchronous motor, provides a 
timing link to the radar antenna. This cam shaft controls the film 
pull-down and ensures that a complete 360 sweep of the PPI tube 
will be photographed irrespective of the antenna speed. 


Figure 8 shows the pull-down actuator. A standard 35-mm 
motion picture intermittent was reworked to pull down five perfora- 
tions. Figure 8 shows the intermittent drive mechanism. A rotary 

Fig. 7. Cycling control system. 

Fig. 8. Film pull-down actuator 



solenoid with a 45 stroke is geared up 8:1 to produce a 360 inter- 
mittent rotation for every stroke. Power is applied to the inter- 
mittent through a ratchet and pawl system. The pull-down is rapid 
(10 millisecond) and accurate. Successive pull-downs register within 
d= in. at a 10-ft image diameter. 

Throughout the design every effort has been made to unitize the 
equipment so that each functional part is easily removable by un- 
skilled field service men. Examples of complete detachable units 
are : the processing head, the control unit, the pull-downs assembly, 
the condenser system, and the objective system. 

In conclusion, we wish to acknowledge the able contributions of 
Mr. Carl Brasser to the design of the unit that has been described. 
It is also a pleasure to express our appreciation for the excellent co- 
operation of the radar development engineers of the Watson Labora- 
tories of the Army Air Force through whose auspices this work has 
been conducted. 

The Trend in Drive- In Theaters 


Summary A brief history pointing to the rapid postwar development of 
drive-in theaters based on fulfilling important needs of a huge "Forgotten 
Audience" is given. An outline of some of the unique design features and of 
the special services made available indicate why drive-in theaters offer an 
economical and versatile means of recreation, relaxation and general enjoy- 
ment for motorists, and a return on capital investment which is an investor's 
dream. Prerequisites for satisfactory projection and sound as basic re- 
quirements for successful drive-in theater construction and management, 
are reviewed. 

THE FIRST DRIVE-IN THEATER was built near Camden, N.J., in 
1933. By the end of World War II, there were only about 60 
drive-in theaters, indicating that the idea had caught on slowly during 
those eight years before the war. It was not until after the war that 
the wave of open-air-see-the-movies-from-your-automobile enter- 
prises really got under way. During the four years since VJ Day over 
1000 have been constructed and many more are being planned or are 
under construction. As soon as unrationed gasoline again became 
available, the public took to the highways for the wide-open spaces. 
The drive-ins were doing capacity business. Prospective theater 
owners could not build indoor theaters at first because of govern- 
ment restrictions on building materials. Then followed prohibitive 
construction costs. It was quickly realized that drive-in theaters 
could be constructed of readily available materials and equipment. 
It was also determined that they could be built at a cost of approxi- 
mately 20% of the postwar costs involved in building an indoor the- 
ater of an equivalent patron capacity, based on an average drive-in 
audience of approximately three persons per car. 

Simultaneously, postwar projection and sound equipments were 
announced which had been designed and built expressly for drive-in 
theater use. It made obsolete most of the equipment used in drive-in 
theaters before the war, especially the sound equipment. The largest 
single factor in contributing to public acceptance of drive-in theaters 
is the in-car speaker, introduced by RCA in 1941 just prior to our 

PRESENTED: June 10, 1949, at the SMPE Central Section Meeting at Toledo, 



entry into World War II. Experience gained previous to the war 
pointed the way to successful drive-in theater construction, equip- 
ment, and management. The in-car speaker removed most of the 
restrictions on locations where the use of centralized speaker systems 
at the screen would have classed them as public nuisances. For the 
first time, a theater patron had complete control over the sound. 

To the amazement of even the drive-in theater owners, in came a 
type of patronage rarely seen at indoor theaters; the physically 
handicapped, invalids, convalescents, the aged, deaf people, expectant 
mothers, parents with infants and small children whole families, 
dressed as they pleased in the privacy and comfort of their own do- 
main on wheels. They are continuing to come in increasing numbers 
from rural, suburban, and city areas a new clientele representing a 
long neglected but highly important segment of some 30,000,000 
people of the "Forgotten Audience," who, according to the claims of 
some producers, had not been attending indoor movie theaters. 
These are the backbone of drive-in theater patronage, and everything 
is being done to retain their acceptance of the drive-in theater. 

Drive-in theater patrons can do as they please within the dictates of 
decency in the privacy of their automobiles. They can shell and eat 
roasted peanuts, smoke, hold a normal conversation, regulate ventila- 
tion, and relax in wider and more comfortable seats with more leg 
room than possible in an indoor theater. There is no parking problem 
or standing in line for admission. Parents are relieved of the worries 
and expense associated with employing suitable baby sitters, or of the 
conduct of their children if left at home. Obviously no drive-in the- 
ater can afford a reputation for being lax in enforcing good conduct. 

Employees like working in drive-in theaters. There is a bit of 
carnival or county fair atmosphere which adds to the spirit of show- 
manship. Even the projectionist finds a difference instead of look- 
ing down toward the screen, he looks up, as do the theater patrons. 


Taking their cues from the gasoline filling stations of the leading oil 
companies, aggressive drive-in theater exhibitors render those extra 
services and courtesies which experience has proven gain public 
favor: windshield wiping, car towing, tire changing, a free gallon of 
gas for dry tanks. Many other services have been made available to 
the public which are customarily not found in. most indoor theaters. 
There are diaper and other vending machines carrying personal 


items, free bottle warmers for baby formulas, a nurse in attendance, 
call service for doctors or others subject to emergency service calls. 

Thus, the drive-in theater has long since passed from the novelty 
category into the realm of big show business. As the number of 
drive-in theaters has increased, picture availability has improved, 
bringing in the regular movie-going public by the car-full. Returns 
on capital investment are an investor's dream and have been so 
startling as to attract new capital from sources far remote from the 
theater business. The maintenance costs of drive-in theaters have 
been estimated to run as low as 20% of those for an indoor theater. 
The concession business of the drive-in theater is the envy of almost any 
roadside stand and is estimated to account for about 25% of the gross 

Each year, rapid strides are manifest in drive-in theaters. There is 
now available a highly scientific modern toll system, a modification of 
collection systems used at the largest bridges and tunnels, which is a 
substantially foolproof method of collecting and recording toll 
receipts, at the same time eliminating the use of tickets. There are 
airplane drive-ins and canoe drive-in theaters. Glamour prevails in 
many drive-in theaters, featuring lighted waterfalls over the rear of 
the screen tower, beautiful landscaping, and ultra modern concession 
stands with exclusive names such as "Snack-N-Vue Bar" and "Din- 
A-Peek Restaurant." In fact, everything is being put into drive-in 
theaters which experience indicates the public likes with their outdoor 
moving picture entertainment. A free rein has been given to the 
imagination in the architects' plans, each new plan vying with those 
of competitors for increased public acceptance of drive-in theater 
environment, offering an economical and versatile means of recrea- 
tion, relaxation, and general personal enjoyment for each and every 

It thus becomes obvious that the selection of a location, the plan- 
ning and the construction of a successful drive-in theater require the 
assistance of an experienced consultant well informed on the many 
complex problems which are involved. Unforeseen costs resulting 
from mistaken judgment on the part of inexperienced builders can 
force undesirable economies in the selection of the most essential ele- 
ments of the over-all enterprise, namely, the projection and sound 
equipment. The prospective investor in a drive-in theater needs to 
be informed as soon as possible that the return on his investment has 
a much better chance of attaining his expectations if the policies of 


good business practice are consistently maintained, rather than an 
attitude of trying to build and equip a drive-in theater as cheaply as 

On the assumption that the drive-in theater has been so planned 
and constructed that each occupant of every parked automobile can 
see the picture on the screen, it follows that the quality of the pro- 
jected picture and reproduced sound is without exception a prime 
requisite for entertainment enjoyment. 

Ever since the first drive-in theater was constructed, the question 
of the amount of light on the screen has been the main bottleneck of 
this type of theater. It has been not too many years since a 30-ft 
screen was considered a large screen for in-door theaters. Today, 
there are a great many drive-in theaters where screens are 60 ft wide 
or larger. The average patron may have the feeling that when we 
double the width of the screen, we should correspondingly double the 
amount of light available. Unfortunately, however, since we are 
talking about screen area, instead of doubling the amount of light we 
have to multiply it by four to retain the same light level over the total 
area of the screen. 

This crying demand for more light on drive-in screens has resulted 
in more powerful arc lamps. In general, as the amount of light at the 
aperture increases, a point is finally reached beyond which it is danger- 
ous to go because of damage to the film. Certain manufacturers have 
introduced heat filters which may remove approximately 40% of the 
heat with a 20% loss in light. Frequently, it is found that excessive 
costs for both carbon and power consumption can be avoided by re- 
ducing the operating amperage and eliminating the heat filter without 
decreasing the light on the screen. In other words, there is no point 
in having a high light level only up to the heat filter and then having a 
20% loss in light, unless the over-all amount of light transmitted to 
the projector is appreciably higher than would be the case if the whole 
setup operated at a lower amperage without a heat filter. 

One other important feature is that a heat filter may have a 20% 
loss in light on the day of installation, but this light loss may appre- 
ciably increase as time goes by, due to two causes: (1) the efficiency 
of the heat filter may decrease with age; and (2) dirt on the surfaces 
of the heat filter will reduce its light transmission. 

Consequently, a heat filter is a unit which is continuously getting 
worse with age. In general, then, very much more effective operation 
can be obtained if a conventional heat filter with its light absorbing 
properties can be omitted. 


There are two general classes of arc lamps currently used in drive-in 
theaters: (1) the reflector type, using amperages of approximately 
80 to 85 amp on a 9-mm black high-intensity positive carbon; and 
(2) condenser type arc lamps, using amperages ranging from 130 to 
180 amp. 

The essential difference between the amplifier systems designed for 
drive-in theater use and those for indoor theater use is the higher audio 
power required for distributing peak signals without distortion to 
large numbers of in-car speakers, often totaling 1000 or more. 

A typical drive-in theater amplifier is shown in Fig. 1. This ampli- 
fier has a total power output of 250 watts. It is a dual-channel sys- 
tem with the inputs connected in parallel, but with the output from 
each channel connected to one-half of the total number of in-car 

At the top of the amplifier rack is the terminal strip for making ex- 
ternal connections. Directly below is the channel selector switch 
and test panel. Next follow the two voltage and two 125- watt power 

Figure 2 shows the manner in which the amplifiers can be turned 
down on their hinges for easy access to the circuits when servicing. 
The channel selector switch makes it possible to operate with both 
channels simultaneously as a dual channel system or, in the event of 
trouble in either channel, to switch the entire speaker load onto the 
output of the operating channel. At the same time, this switch 
changes the output transformer tap to match the speaker load. The 
disabled amplifier is automatically disconnected from the a-c power 
source and the output load, so that it can be repaired without inter- 
rupting the performance. Monitoring and testing facilities are also 
included on the selector switch panel. 

The voltage amplifiers are two-stage units having high impedance 
inputs and transformer coupled outputs. The soundheads are con- 
nected to the inputs by means of low-capacity cables. Coupling 
between the voltage and power amplifiers is accomplished through a 
500-ohm "H" type variable attenuator which serves as the volume 
control. The attenuator is connected as a dual 250-ohm variable "T" 
attenuator so that both channels are equally controlled. 

The power amplifiers are three-stage Class "B" units utilizing four 
809 type tubes in the output stage and are rated at a 125- watt output 
each with less than 3.5% distortion between 50 and 5000 cycles. 
Approximately 10 db feedback between the output and driver stages 


.,. ; " 

Fig. 1. Drive-in theater amplifier. 

Fig. 2. Amplifiers turned down. 





contributes to holding the distortion to this \o\\ value, and also serves 
to hold the output level substantially constant with variations in 
speaker load. 

Figure 3 is a ramp station comprising two speakers and a junction 
box. The speaker housings are of die-cast aluminum, rugged enough 
to withstand being run over by an automobile without crushing. 
They are small in size and light in weight and are easily handled with 
one hand. 

Fig. 3. Ramp station of two speakers and junction box. 

The hook or neck construction was designed so the speaker can be 
hung on the car window, with the window almost closed, as would be 
necessary in rainy or cold weather. 

The volume control knob is of bright red plastic and is tamper- 

A simple rheostat volume control is used in the voice coil circuit of 
the speaker. 

The mechanisms used in these speakers are especially designed for 
drive-in speaker use. All metal parts, including the magnets, are 
heavily plated with cadmium. The magnets are anchored to the 


frame so that they cannot shift and cause the pole piece to move off 
center. The voice coil and diaphragm are waterproofed and con- 
structed to withstand all outdoor weather conditions, including floods. 
A drive-in theater at Endwell, N.Y., was under water for three days, 
submerging all of the speakers and junction boxes. When the theater 
was finally reopened, all but three of the speakers played perfectly. 

The junction boxes are also of die-cast aluminum and have the same 
type of finish as the speakers. 

Figure 4 shows the post and road light, and the method used to 
obtain a cone of light at the base of the post and project an adjustable 

Fig. 4. Post and road light. 

beam of light out into the driveway in the shadow area between the 
rows of parked cars. The junction boxes are available with or with- 
out this feature. 

The miniature 28-volt, .17-amp lamp is supplied by a 32- volt power 
transformer located in the booth, with the voltage dropped through 
specified line resistors. The lamp was designed for airplane use and 
has a rugged shockproof filament assuring long life. This type of 
lighting eliminates any apprehension on the part of the automobile 
driver who is obliged to turn off his headlights on driving into 
the theater. There can be no fear of hitting an unseen person or 
object, because each roadway light serves continuously to usher the 


DRIVE-IN Tin \ i 


driver safely toward a parking space. The ]>;ii!,<in of elongated 
lighted areas in the roadway follows the curved contour of each ramp. 
The over-all effect, including the lighted areas at the base of each post, 
gives the drive-in theater a beautiful appearance, with ample lighting 
within the parking area at a minimum of cost. In a 1000-car theater, 

Fig. 5. Junction box with lamp for concession attendant. 

Fig. 6. Mechanical device for concession signaling. 

for instance, there are 1000 beams of light from 500 tiny lamps with a 
total power consumption of only approximately 2 kw. 

The transformer is completely moisture proof. It has a high im- 
pedance primary winding which permits connecting many in parallel 
across the amplifier output. 


For concession service, an additional miniature lamp is installed in 
the junction box which is readily seen by the concession attendant as 
a red glow from a lens located in the junction box cover, Fig. 5. The 
patron controls the light by means of a toggle switch installed in the 
face of the speaker housing. 

Three types of speaker cables take care of basic requirements. 
There is the low cost straight cord, and the deluxe Koiled Kord. A 
theft resisting cable includes a hardened stranded steel cord which is 
anchored at both ends. It also includes three conductors and is 
standard for use with electrical concession lights. 

Another type of concession signaling device is entirely mechanical, 
Fig. 6. It is simply a stainless steel slide attached to the under part 
of the junction box base, and so constructed that the patron can push 
or pull it and cause a red lens to intercept the down light beam of the 
post light, causing it to change from white to red. The lens is plainly 
visible to the concession attendant. 


The problems of daylight projection involved in endeavoring to 
obtain longer daily operating hours, and of in-car heating so as to 
expand the operating seasons, are apparently being given much 
thought. They both need a practical solution applicable to every 
section of the country. Fog is a serious problem in some areas, 
occasionally becoming heavy enough to cause refunding of admis- 

Regardless of such remaining problems, there is every indication 
that the public needs and wants more drive-in theaters, if strategically 
located, wisely constructed, and properly equipped. 

The trend is toward drive-in theaters having smaller car capacities 
which can adequately serve rural or suburban communities. Many 
have already outgrown their car capacities. The solution has been 
simple and economical in those theaters owning sufficient land. It 
has been necessary only to add and equip one or more ramps. 

Indications are that an undetermined number of well-established 
drive-in theaters which have been in operation for several years have 
made plans for improvements and for replacing their old equipment. 

All of these activities are conclusive proof that the drive-in theater 
business is here to stay. Exhibitors were literally pushed into it as an 
aftermath of World War II. In the opinion of the author, only a 
World War III can be its Nemesis. 

A Sturdy, High-Quality 
16-Mm Projector 





Summary A 16-mm sound motion picture projection system according to 
U.S. Navy Bureau of Ships Specification CS-P-41A must meet high-quality 
standards for picture and sound reproduction, while retaining the necessary 
durability for Navy use. It must provide high light output with good light 
distribution, long film life, low flutter, low distortion, and good frequency re- 
sponse. A projector designed for this purpose uses a sprocket type intermit- 
tent and unit component construction for ease of maintenance. The paper 
describes the projector mechanisms and amplifier circuits, and states per- 
formance results. 

described has been developed in part under a contract with the 
United States Navy Bureau of Ships. The specification which has 
been used as a guide throughout this development is in many respects 
an extension of Specifications American Standard Z52.1 and Joint 
Army-Navy P-49. In effect, it calls for a system which has per- 
formance approaching 35-mm standards, as well as the necessary 
sturdiness to withstand severe Navy use. Initial review of conven- 
tional 16-mm components indicated that one or the other of these 
requirements would have to be sacrificed to some extent if conven- 
tional 16-mm techniques were followed throughout. The most 
promising approach seemed to be to incorporate the desirable features 
of 35-mm techniques into a 16-mm projector. In keeping with this 
idea, new components were developed where necessary and were 
adapted to existing components to establish the projector design. 
The amplifier and loudspeaker units were developed with similar ob- 
jectives in mind. 

The three basic units, projector, amplifier and loudspeaker, may be 
operated as a single-projector, single-amplifier system (Fig. 1), or 
may be expanded into any combination of two projectors and two 

PRESENTED: April 7, 1949, at the SMPE Convention in New York. 





amplifiers. For change-over operation, a junction box is required 
to interconnect the two projectors. Each unit is housed in a separate 
carrying case which has space for accessories and operating spares. 
The cases are made of a newly developed sandwich laminate of cellu- 
lar-cellulose-acetate bonded between thin aluminum sheets. Panels 
of this material are lightweight, but are extremely rigid. Covers are 
gasketed to seal against moisture and dust. 

The projector mechanism is shock-mounted at four points in the 
carrying case. Sealed ball bearings lubricated with high temperature 

Fig. 1. Projector and amplifier. 

grease are used on all shafts except those in the intermittent mecha- 
nism, where sleeve bearings are used. To minimize the effect of 
corrosion, the materials employed are protected aluminum alloys, 
copper alloys, corrosion-resistant steel, or fiber. 


The projector (Figs. 2 and 3) operates from 115- volt 60-cycle alter- 
nating current. Film is transported at the sound speed only, namely, 




Fig. 2. Projector, mechanism side, cover removed. 

Fig. 3. Projector, operating side. 


24 frames/sec. Reel arms, tilt mechanism, and electrical outlets are 
wholly enclosed within the carrying case, and a 2000-ft reel is carried 
in the cover of the case. The reel arms are an integral part of the 
projector, and are both positioned above it in operation. For ease 
in assembly and service, the projector uses a vertical main frame to 
which components and subassemblies are fastened. The components 
which make up the projector are discussed below. 

Drive Motor 

An 1800-rpm, synchronous type drive motor is used in preference 
to a governor-controlled unit. Although the synchronous motor is 
the larger and heavier of the two, its quieter operation, more constant 
speed, and the fact that it requires no brushes are decisive points in its 

The centrifugal starting switch in the motor is interconnected with 
the projector controls so that the projection lamp cannot be energized 
unless the motor is running. 

Intermittent Mechanism 

The projector uses a sprocket type intermittent mechanism which 
provides a four-tooth continuous film control. This type of inter- 
mittent provides a favorable stress distribution at the sprocket holes, 
and has a corresponding advantage in regard to film life. Film life 
tests 1 have shown that several thousand passages may be made with- 
out damage to the sprocket perforations. 

The mechanism is housed in an oil-filled gear box receiving an input 
drive from the synchronous motor and providing output drives for 
the shutter, intermittent sprocket, and the vertical shaft. The shut- 
ter shaft is driven at 1440 rpm through a single mesh of spiral gears. 

The principal parts of the mechanism are shown in Fig. 4. The 
12-tooth intermittent sprocket (1) is indexed by a 12-tooth star 
wheel (2) and a cam (3) which has a 50 operating section on the 
shutter shaft. The star wheel teeth are held in contact with the 
dwell surface of the cam by a preloaded, spirally coiled, flat spring (4). 
The center of the spring is fixed to the star wheel shaft and the outside 
of the spring is fastened to a spur gear. The spur gear unwinds the 
outside of the spring at a uniform rate while the center is being wound 
up at the same average rate by the intermittent action. Using spring 
loading in this manner not only simplifies the cam construction but 




also provides automatic wear take-up during the life of the mecha- 
nism. Since but one surface is required lor tooth registration, ma- 
chining of a tooth-confining cam surface and the need for dose t ooi h- 
thickness tolerances are eliminated. 

Framing is accomplished by moving the shutter shaft axially. 
This movement shifts the registration surface of the cam, rotating 
the intermittent sprocket. The projection aperture thus remains 
stationary on the screen during framing, and the picture moves wit hin 
this aperture. 

Sealing the gear box against oil leakage has been effected by the 
use of Johns-Man ville synthetic rubber shaft seals which are pressed 
into the housing and flexibly grip the shafts. 


CAM (3) 

Fig. 4. Schematic of intermittent mechanism. 

The shafting arrangement of the intermittent mechanism has per- 
mitted the use of a large disk type shutter to achieve a relatively high 
light efficiency. With an 8^-in. diameter shutter, located ? in. 
from the film plane, a light efficiency of 73% has been attained. 

A V-belt pulley assembled as part of the shutter drives a 3^ie-in. 
diameter blower- wheel at 2450 rpm for ventilating the lamp house. 
The relatively slow operating speed minimizes blower noise. How- 
ever, the blower wheel is large enough to realize a free-air discharge 
rating of 140 cfm at this speed. 

Gate and Film Trap 

The gate and film trap have been designed to register the film on 
the emulsion surface for standard 16-mm emulsion position, and to 
guide the film from the sound track side. Guiding the film in this 
manner conforms to ASA specifications. Chandler, Lyman, and 
Martin 2 have ably discussed the film guiding problem with respect to 


lateral film shrinkage. When the film is guided at the perforated 
edge, a rail between sound track and picture area which will not 
encroach on the sound track becomes impracticably small. This 
results from the fact that the sound track, being opposite the guided 
edge, will be shifted nearly the full amount of the shrinkage. In 
guiding the film from the sound track side, the shift between the 
picture and sound due to shrinkage is a smaller proportion of the 
total, and therefore a reasonably large rail can be provided. Regis- 
tration on the emulsion side of the film eliminates any effect of in-and- 
out-of focus which might arise from variation in film base thickness. 

For threading, a lever at the top of the film trap casting retracts 
the spring-loaded, film side guides and, at the same time, lowers 
the intermittent sprocket shoe. The film trap body and pressure 
shoe form a funnel-like entrance which defines the film plane. The 
film, in entering, depresses the pressure shoe, providing the necessary 
gate opening. The film trap casting is removable for cleaning pur- 
poses, at which time the pressure shoe also becomes accessible. 

Lamp House 

The complete lamp house assembly is mounted at the back of the 
main frame. The projection lamp and optical elements are supported 
on the lamp house door, and are thermally insulated from it. Either 
a 750-watt or a 1000- watt incandescent lamp may be used. 

The projection lamp is a newly developed General Electric type 
which burns base up. It has a medium ring, double contact base. 
The mechanical mounting ring on the base is independent of the elec- 
trical contacts, and is large enough to permit very accurate position- 
ing. Filament size, construction, and bulb size are consistent with 
current practices. However, since the lamp operates with the fila- 
ment in the downward position, the cooling air strikes the hottest 
portion of the lamp first, which should result in more effective cooling 
and correspondingly longer lamp life. Also, any blackening of the 
interior wall due to evaporating tungsten occurs far enough above the 
optical axis so that there is very little loss of light during the life of 
the lamp. 

When the lamp house door is open, the lamp is easily lifted out for 
replacement; it is held in position by the spring-loaded electrical 
contacts only when the door is closed. Opening the door actuates a 
safety switch which de^energizes the contacts in the top of the lamp 

1950 16-MM PROJECTOR 177 

The projection optics consist of a Bausch & Lomb 22-mm single 
aspheric condenser lens with a 39^-mm spherical reflector, positioned 
behind the projection lamp. The combination of an accurately 
positioned projection lamp and these optics produces the high screen 
illumination and good light distribution shown in Table I. 



Light Output (shutter running), 750-watt lamp: 320 lumens 

1000-wattlamp: 430 lumens 

Uniformity of Illumination (% of center), average corner: 92% 

lowest corner: 85% 

Picture Unsteadiness (% of picture width), vertical: 0.2% 

horizontal: 0.1% 

Flutter (measured on RCA Meter No. MI-9763-B), average: 0.25% 

maximum: 0.35% 

Noise of Projector Mechanism: 63 db above 10 ~ 10 watts /sq cm 
Power Output: 20 watts with less than 2% distortion from 100-4000 c/sec 
Frequency Response from Film: 3 db below midband level at 80 and 5000 c/sec 
Electrical Noise Level (controls set at normal): 55 db below 20 watts 

Lens Mount 

The lens mount accommodates objectives having the dimensions 
shown in JAN-P-49. A one-piece focusing sleeve, cut longitudinally, 
clamps and locates the lens barrel. The design utilizes the elastic 
properties of the sleeve to grip the lens barrel securely, while per- 
mitting ready release by a cam crank in the longitudinal cut. The 
end of the sleeve forms an accurate register surface for the shoulder 
on the lens barrel, so that the lens can be returned to exact position 
after removal for cleaning. A double thread on the outside of the 
sleeve provides fine focusing. 

The projector is equipped with a Bausch & Lomb 2-in., //1. 6 
Supercinephor "16" lens which resolves better than 90 lines/mm over 
the whole screen. Lenses with focal lengths between two and four 
inches can be accommodated. 


The take-up device utilizes the weight of the film on the reel to 
provide variation in driving torque so that reasonable film tension 
may be maintained under all conditions. With proper adjustment, 
a 5-oz tension in the film may be obtained with an empty reel, and 


this tension decreases to approximately 2 oz when 2000 ft of film 
have been taken up. 

Rewinding is accomplished without interchanging reels. The re- 
wind drive, engaged by a mechanical switch near the feed sprocket, 
handles 2000 ft of film in 3% mm. A slip clutch allows the full reel 
to coast to a stop without damage to the drive when the projector is 
shut off. 

Sound Head 

The sound scanning mechanism is an integral unit assembled 
through vibration isolation mounts to the main frame. The pre- 
focused exciter lamp is operated from a radio-frequency power supply 
in the amplifier. The Bausch & Lomb scanning lens tube is similar 
to those used in 35-mm projectors, except that the adjusted-jaw me- 
chanical slit has been reduced in size so that the width of the slit image 
is approximately 0.0005 in. The scanning beam thus obtained is 
independent of the size of the lamp filament. From the theoretical 
consideration of the scanning losses 3 the zero-output frequency for 
an 0.0005-in. scanning beam is 14,400 c/sec (cycles per second); 
at 7,200 c/sec, the output is down only 4 db. 

A double convex lens mirror focuses the light beam onto the cathode 
of a blue-sensitive gas photocell. For film having emulsion in the 
nonstandard position, a plane parallel piece of slide cover glass ap- 
proximately 0.01 1-in. thick is inserted in the light beam to shift the 
focus from one surface of the film to the other. In this way the fine- 
ness of the slit image is retained for film with nonstandard emulsion 
position. Provision has been made for focusing the optic tube as well 
as adjusting it in azimuth and in lateral position. 

The mechanical filter 4 which transports the film through the sound 
head consists of a film-driven eddy current drag sprocket, a flywheel 
on the scanning drum shaft, and a flexibly driven sound pulling 
sprocket. The inertia of the flywheel on the scanning drum shaft 
and the compliance of the spring in the flexibly driven sound sprocket 
determine the low cutoff frequency of the filter. The eddy current 
drag sprocket damps any oscillation at the natural frequency of the 
filter, and also provides frictional torque for rotating the scanning 
drum so that a pad roller with its attendant difficulties is not re- 
quired. The coil in the flexibly driven sound sprocket is preset in 
relation to the torque imposed by the eddy current drag so that a film 
loop is formed on starting the projector. The film loop thus formed 

1950 K)-MM PROJECTOR 17<> 

isolates the intermittent from the sound head and at the same time 
establishes synchronization between picture and sound. A second 
film loop between sound and holdback sprockets isolates the sound 
head from the effect of the take-up. 


The amplifier has been designed to meet Navy mechanical and 
electrical specifications, and consequently differs considerably from 
amplifiers used in current commercial projectors. It supplies a power 
output of 20 watts to a 4-ohm load, with distortion considerably below 
the required maxima of 2% between 100 and 2,000 c/sec, and 4% 
between 2,000 and 4,000 c/sec. The normal frequency response of tho 
amplifier is 3 db below the midband output level at 80 and 8,000 
c/sec. Separate treble and bass tone controls provide adequate range 
to compensate for most acoustic and film deficiencies. The amplifier 
gain is sufficient to develop a 20-watt output from 400-c/sec level 
test film, with a reserve gain of 20 db when minimum specification 
limit tubes are used. Connections are provided for microphone input 
and for a monitor speaker. 


The circuit consists essentially of an input feedback loop, one stage 
of amplification containing the volume and tone control circuits, and 
an output feedback loop. 

A pentode and a triode stage are included in the input feedback 
loop where about 25 db of over-all negative current feedback are 
used. The feedback reduces the effective input impedance of the 
amplifier to the extent that a capacity as high as 200 micromicrofarads 
in the photocell cable causes no appreciable change in the amplifier 
output at 10,000 c/sec. A 15-ft length of RG71/U signal cable could 
thus be accommodated. Cable microphonics, hum, and distortion are 
also attenuated. 

A 40-db attenuator having 2-db steps is used as a volume control. 
From this control, the signal is amplified in a single triode stage which 
drives the tone control network. The range of boost and attenuation 
afforded by the separate low and high frequency controls is slum n in 
Fig. 5. 

The output feedback loop includes three stages : a triode amplifier, 
a triode phase inverter, and a power stage using push-pull 6L6GA 
tubes in class ABi operation. Feedback voltage developed across the 




secondary of the output transformer is coupled to the cathode of the 
triode amplifier stage, giving 10 db of over-all feedback. 

Exciter Lamp Supply 

An oscillator type power supply for the 6.5-volt, 2.75-amp exciter 
lamp is built into the amplifier chassis. The oscillator uses a single 
6L6GA tube which operates at a frequency of 30 kilocycles per 
second, and has an efficiency of about 50%. With this type of power 
supply, the exciter lamp filament is not subject to cyclic temperature 
variations within the audio range, as it would be if it were operated 
from a 60-cycle source. 






100 1000 


Fig. 5. Frequency response. 

10000 20000 


The components have been selected to give as reliable operation 
as possible under extremes of shock, vibration, temperature and hu- 
midity. All paper condensers and transformers are hermetically 
sealed. The carbon tone control potentiometers are of the molded 
element type. Insulated carbon resistors and vitreous enameled 
power resistors are used throughout, and are operated at less than 
50% of their power ratings. Three hermetically sealed plug-in 
electrolytic condensers are required as cathode by-pass units, but the 
amplifier will function at slightly reduced gain with any of these 
removed. All other condensers are either mica or oil-filled paper 
units, operated at less than 75% of their voltage ratings. 

1950 16-MM PROJECTOR 181 

The aluminum alloy chassis is givm ;i Iii-Ii device of rigidity by 
partitions and angle reinforcing members which also provide shielding 
and mounting supports for terminal strips. The chassis is secured in 
the carrying case by a shock absorbing aircraft type rack, and may 
be removed for servicing by loosening two thumb-nuts. Spare tubes, 
electrolytic condensers and lamps are housed in the amplifier carrying 

Fig. 6. Loudspeaker. 


The Western Electric Model 754B loudspeaker, used with this 
equipment, is a 12-in. diameter unit with a 4-ohm nominal impedance. 
It has a high power handling capacity, and is constructed to with- 
stand heavy shock and vibration. It is the only known commercially 
available reproducer having both the required efficiency and a 
phenolic impregnated cloth cone. This loudspeaker is operated with 
the back of the carrying case (Fig. 6) closed, the enclosed volume being 
approximately that recommended by the manufacturer. 



A development of this nature represents necessarily the combined 
efforts of many individuals. The authors gratefully acknowledge 
the collaboration of Mr. E. G. Mercier and Mr. A. F. Hayek. Thanks 
and appreciation are also due those who have assisted in the various 
stages of the work. 


(1) C. F. Vilbrandt, "The projection life of 16-mm film," Jour. SMPE, vol. 48, 
pp. 521-542; June, 1947. 

(2) J. S. Chandler, D. F. Lyman, and L. R. Martin, "Proposals for 16-mm and 
8-mm sprocket standards," Jour. SMPE, vol. 48, pp. 483-520; June, 1947. 

(3) N. R. Stryker, "Scanning losses in reproduction," Jour. SMPE, vol. 15, 
pp. 610-623; November, 1930. 

(4) W. J. Albersheim and D. MacKenzie, "Analysis of sound film drives," 
Jour. SMPE, vol. 37, pp. 452-479; November, 1941. 

Animar Series of Photographic Lenses 


SummaryA new series of highly corrected lenses, with focal lengths 
ranging from 12.7 mm to 100 mm, has been designed for 8-mm and 16-mm 
cinematography. The speed of some lenses in this series is as high as //1. 5. 
Some general problems of optical design and of evaluation of lens per- 
formance are discussed. Strict criteria are used for appraising the quality 
of these lenses, which will be introduced on the market under the name of 

ABOUT three and a half years ago the late Dr. W. B. Ray ton intro- 
duced 1 to the Hollywood convention of this Society a series of 
new//2.3 Baltar lenses of short focal lengths. These lenses were pri- 
marily intended for new professional 16-mm cameras that were reaching 
the final design stages and were about to be introduced on the market 
at that time. Considering the preference of some camera manufac- 
turers for using their own focusing mounts, and because of uncertain- 
ties as to whether all camera manufacturers would eventually adhere 
to the same mounting dimensions, the Baltar lenses were offered 
in plain barrels, and the fitting into focusing mounts for any particular 
camera was left to the camera manufacturers. This has limited the 
use of Baltars to just a few types of cameras whose manufacturers 
have been willing to undertake the task of mounting. 

In the meanwhile the extensive engineering efforts of many indi- 
viduals and organizations have resulted in a number not only of 16-mm 
professional, but also of 16-mm semi-professional and 16-mm and 
8-mm amateur cameras. Most of these 16-mm cameras are capable 
of accommodating lenses having the mounting thread of 1.000 in. 
X 32 and the registration distance of 0.690 in., and practically all 
8-mm cameras accommodate lenses of 0.625 in. X 32 thread and the 
registration distance of 0.484 in. Thus a wide interchangeability of 
lenses now is possible and an extensive market is available for lenses 
of these standard mounting dimensions. 

Characteristic of this market is the demand for high-quality 
products at a reasonable price. Due to the great improvements made 

PRESENTED: April 8, 1949, at the SMPE Convention at New York. 



during and after the war in engineering and manufacturing, camera 
manufacturers have been capable of meeting the quality requirements 
to such an extent that in many cases the difference between a "profes- 
sional" and a "nonprofessional" camera can be attributed riot to some 
performance deficiency of the latter but only to its lacking a number of 
special features which are considered essential for efficient motion 
picture production on the Hollywood level. Disregarding some ex- 
ceptions, it seems fair to state that the basic limitations of the present- 
day 16-mm and 8-mm motion picture photography lie not in the 
mechanical shortcomings of cameras but in the inherent restrictions 
imposed by the laws of optical image formation (problems of depth 
of field and difficulties of exact focusing with lenses of short focal 


For 8-mm Cameras For 16-mm Cameras 

Focal Diagonal Focal Diagonal 



coverage, * 



coverage, * 







All these lenses have the standard mounting threads and the registration dis- 
tances mentioned in the beginning of this paper. 

The lenses for 16-mm cameras may be mounted on 8-mm cameras by means 
of special adapters. 

* Based on the projector frame. 

f Named Tele-Animars. 

lengths) and by the materials used for photographic recordings 
(granularity of emulsion and warpage of film). Some of these optical 
factors will be evaluated in another paper planned for a future 

The mechanical excellence of many postwar cameras and the gen- 
eral education of the public to the appreciation of better motion 
picture photography necessarily impose the requirement of high 
quality also on the lenses for these cameras. This requirement is 
eminently met by the Baltar series. However, the requirement of a 
price which "the general public can afford" and the occasional need 
for lenses faster than //2.3 Baltars had to be met by another series 


of lenses, which are being introduced on the market under the name 
of Animar. A list of these lenses is given in Table I. 


The purpose of designing this series was to produce lenses that would 
be economical from the point of view of the manufacturer and the 
purchaser. Hence an effort was made to utilize a minimum number 
of components without sacrificing the basic requirement of high- 
quality performance. Extensive research and development work 
was needed for meeting this condition. Finally it was found possible 
to use not more than four elements for practically the whole series, 
with the exception of the //1. 5 lenses in which design six elements 
were necessary in order to meet the desired quality of performance. 

In this connection and for a better appraisal of the results, it seems 
worth while to discuss some misconceptions with regard to design and 
performance evaluation of optical systems. These misconceptions 
are: that a high quality of optical performance cannot be achieved 
unless many elements are utilized in the lens construction; that there 
are some new developments in lens design which make obsolete all 
previously used formulas; that a formula originally developed for a 
certain angular coverage can always be improved if the required 
angular coverage is reduced; and that resolution tests can serve for 
absolute evaluating of lens quality. All these statements have grains 
of scientific truth, but they cannot be accepted without severe reserva- 

There is no general rule for determining the minimum number of 
elements necessary for producing a high-quality optical system. For 
example, in astronomical telescopes, which require almost perfection 
in performance, only two elements may be sufficient because of the 
extremely narrow angular coverage at a low speed. More than two 
elements are usually required for covering an extended field at a 
speed higher than about //8. It was a truly great contribution to the 
field of photographic optics when in 1893 H. D. Taylor developed a 
triplet design (frequently known as the "Cooke Lens" because at that 
time he was with Cooke and Sons of York, England), and clearly 
demonstrated that three elements may be sufficient for obtaining a 
speed as high as //4 and a coverage up to at least 25 from the axis. 
The suitability of triplets has been later extended to speeds approach- 
ing //2.5 Disregarding some special designs such as the Kellner 2 and 


Schmidt systems employing reflective components and correcting 
plates, it seems that at least four elements are needed for obtaining 
an extended coverage at speeds in the neighborhood of //2.0, or even 
higher if the coverage is limited to a few degrees from the axis. One 
of the principles of fast four-element designs was discovered by J. 
Petzval as early as 1840, and ever since it has been employed for 
producing very satisfactory photographic and projection lenses. 
The Petzval constructions frequently fail, however, when the require- 
ment is to cover a field greater than about six degrees from the axis 
at a high speed ; then more elements are usually necessary for a satis- 
factory image quality in the periphery of the field. 

It is fortunate that mostly moderate angular coverages are involved 
in motion picture photography. Thus with "standard lenses" (focal 
length of K in. for 8-mm cameras, and of 1 in. for 16-mm cameras) 
the diagonal field actually utilized for projection is less than 27 
(that is less than 13.5 from the axis), and it may be just a few degrees 
for a "telephoto" lens. Therefore, three- and four-element designs 
may be particularly suitable in this application. 

In regard to recent or future new developments in lens design, it 
should be sufficient to state that a thorough knowledge of fundamental 
principles and limitations of optical systems had been acquired by 
the end of the last century, and that practically all basic constructions 
were already in existence in the early part of this century. The 
progress since that time has not revealed itself in any fundamental 
discoveries, but in a broader exploration and a deeper understanding 
of theoretical relationships pertaining to optical design, and in their 
practical application to further development of some promising rudi- 
mentary forms. By extreme modifications and extensive compound- 
ing of these forms, some designers have succeeded in exploiting the 
inherent, but previously overlooked or not explored, possibilities, and 
in developing novel constructions of almost spectacular character- 
istics, thus practically reaching the limits of the possible in optical 
design. The evolution of the Gaussian type objectives, mentioned 
later in the text, is particularly illustrative of the creative thinking 
in lens design. One of the major factors in this progress has been the 
increased choice of optical glasses. There is, however, no hope that 
this choice may be extended indefinitely, and no hope that some new 
revolutionary principles of design will be discovered. It seems that 
the time is approaching when major explorations in optical design 
should come to an end with nothing left to do but the tedious task 


of cleaning up details, which is, according in a recent paper l>y ( ieor^e 
Gamow, 3 a fate looming for physics in general. Ad ually t he 'clean- 
ing up of details" has, been, for some time, tin- major activity ot opti- 
cal designers. This activity has resulted in many excellent system- 
providing either speeds or angular coverages or both tar above those 
that could be anticipated with the basic forms. The successful 
development of these systems, however, only emphasizes the fact that 
unless some radically simplified solutions of design problems are dis- 
covered, no great advances should be anticipated either from further 
modifications of the now available forms or from increasing their 
complexity by further compounding. 4 As the matter stands now, it 
is highly illusory to entertain hopes that the desired simplified solu- 
tions are forthcoming. 

It is superfluous to state that, if the best possible results are to be 
achieved, an optical system should always be designed with due regard 
to the intended application. This does not mean, however, that a 
lens designed for a given coverage can be readily made better for a 
less extended coverage, or that some other system can be chosen for a 
superior performance within the reduced field. The fact is that each 
case should be treated in accordance with its merits, and that quite 
often no substantial gains can be made by redesigning a system for a 
narrower coverage. This is particularly true when the original system 
was designed for a relatively small coverage, and its oblique aberra- 
tions were reduced to a negligible minimum. Then the performance 
of the system within a still smaller coverage will be determined pri- 
marily by its residual spherical aberration. If the residual in the origi- 
nal design was reduced to a minimum desired by the designer, it 
would be purposeless for him to attempt a redesign of this system for 
a smaller coverage. 

In the course of this work the goal was to produce a series of lenses 
whose aberrations would be most favorably balanced for the field 
of the 8-mm motion picture frame, and to produce another series 
specifically for the 16-mm frame. Still all the lenses intended for 
16-mm cameras will perform equally well on 8-mm cameras. If, for 
example, a question should be raised whether or not the design of the 
100 mm //3.5 Tele-Animar could be advantageously modified for 
8-mm coverage, the answer would be an emphatic "no," because the 
residual aberrations of this form were reduced to insignificant amounts 
even for a larger coverage, that is for the entire area of the 16-mm 
frame. As a matter of fact the correction of this formula is so satis- 


factory that it can be used for covering the 35-mm frame without sub- 
stantial sacrifices of the image quality in the extended field, excepting, 
perhaps, its very margin. 

The problem of determining the performance of a lens has often 
been reduced to a recording of "its photographic resolution" on the 
assumption that a resolution record can reveal the intrinsic quality 
of the lens. Testing for resolution is of much value in evaluation of 
lens performance; nevertheless, the method has definite limitations 
which have been recently surveyed by a number of investigators in- 
cluding one of the authors. 5 - 6 The basic fact is that a photographi- 
cally recorded resolution pattern generally reveals neither the per- 
formance of the lens itself nor that of the emulsion itself, but that 
of the lens-emulsion combination. 

If the lens is practically free from aberrations, its intrinsic resolution 
potentialities are indicated by its Airy disk (discussed in the papers 
referred to above) whose diameter is a function of the lens //number 
and of the wavelength of light used in the image formation. The 
photographic resolution obtainable with a nearly perfect lens is 
essentially limited by the resolving power of the emulsion used, unless 
the emulsion resolving power is higher than the "Airy resolution" 
of the lens; then the lens becomes the limiting factor. 7 

The situation becomes much more complicated in the presence of 
aberrations. Then the lens itself becomes the primary limiting factor 
which always tends to keep the attainable photographic resolution 
of a lens-emulsion combination under the maximum resolution of the 
emulsion. Therefore, if a lens within a certain area of its coverage is 
capable, for example, of recording a resolved pattern of 100 lines/ 
mm on some emulsion (whose resolving power will have to be some- 
what higher), no conclusion can be drawn that the "resolving power 
of the lens" will not be the limiting factor in photography on emul- 
sions of resolving power lower than 100 lines/mm. Indeed it may 
happen under some conditions that in photography with the same 
lens on a low-resolution emulsion the resolution will be significantly 
below the rated resolution of the emulsion. 

Due to the extreme complexity of the situation and the lack of 
standardization of resolution tests, data on the resolution recorded 
with a lens may be highly misleading if they are not accompanied by 
an identification of the target and the emulsion used in the tests, and 
if they do not reveal the lens performance within the entire intended 
area of coverage. Even when these qualifications are met, the data 


may be of little value in absence of comparative data for similar 
lenses of other makes. Such data are generally not available. 

Attempts have been made to determine the inherent performance 
characteristics of image-forming systems either by an analysis of 
the energy distribution within aberrated image patterns (D. G. 
Hawkins and E. H. Linfoot, 8 M. Herzberger, 9 A. Marshal, 10 and 
others) or by the ingenious recording of response-resolution curves 
of the systems (O. Schade 11 ). These investigations are definitely a 
move in the proper direction. Their application to routine rating 
of photographic lenses will have to wait, however, until the methods 
and the instrumentation involved become more widely accepted 
among optical industrial laboratories, and until an extended experi- 
mental material is accumulated. 

Irrespective of the significance that may be attached to these inves- 
tigations or to any resolution data, the residual aberrations of a lens 
are the primary indicators of the degree of perfection attained in its 
design. The situation here is also very complex as, in order to obtain 
a satisfactory performance, it is necessary not only to reduce all the 
aberrations to an acceptable minimum but also to obtain a most favor- 
able balance of all the residuals within the entire field of coverage. 
Because of this complexity and of the unavoidable compromises, 
it is impossible to rate a state of correction of a lens by some numeri- 
cal coefficient that would express the relative standing of a design 
among other designs. Nevertheless, the rather extended published 
material on the correction of various formulas provides a relatively 
good basis for a qualitative comparison, and, in the case of a high 
degree of correction, the comparison may be based on the tolerances 
derived from the Rayleigh quarter-wavelength criterion of perfection. 
The basic forms of these tolerances may be found in the well-known 
book by Conrady. 12 These tolerances will be used for an appraisal 
of the representative Animar formulas. 

It should be understood that, when used in connection with photo- 
graphic optics, the Rayleigh tolerances seem so severe that optical 
designers generally do not hesitate to accept residual aberrations 
several times greater than would be acceptable on the basis of the 
Rayleigh criterion. The problem of meeting these tolerances be- 
comes increasingly difficult for lenses of longer focal lengths because 
the linear aberrations in lenses derived from one basic formula are 
directly proportional to the focal length, while the Rayleigh limits are 
independent of the focal length. Considering these facts, Conrady, 




a man of unquestionable authority in optical design, found sufficient 
justifications for establishing less strict tolerances for coma, astigma- 
tism, and curvature of field. In the appraisal of the Animar series 
these practical tolerances will be referred to, as well as the strict 
tolerances based on the original Rayleigh criterion. 


The Animar series, in its present stage of development, consists 
of lenses utilizing three, four, and six elements in their construction. 

Since the first triplet was derived by H. D. Taylor, many successful 
efforts have been made to improve the formula and extend its useful- 
ness. There are now a multitude of triplet formulas recorded in 
general and patent optical literature, and many triplet constructions 
have been widely utilized by manufacturers of photographic and pro- 
jection lenses. A number of triplet formulas have been continually 

Fig. 1. Basic form of triplet Animars, 
U.S. Patent No. 2,453,260. 

//2.7, Focal Length 100 mm 

Lens I Ri = 

R 2 = 

II R 3 = 

R 4 = 


55.65 D 2 = 
39.75 L 2 = 
L 3 = 

III R 5 = 107.56 D 3 = 

8.74 ND = 1.6170 

11.05 V = 55.0 

2.78 No = 1.6490 

7.63 V = 33.8 

R 6 

9.54 ND 

= 1.6170 
= 55.0 

used by Bausch & Lomb. Some of them could be readily and satis- 
factorily employed for this particular application. Still, it was con- 
sidered desirable to explore further possibilities of a better design, 
and extensive studies were made of the existing triplet constructions 
before the basic formula was finally derived for the triplet Animars. 13 
This basic formula is represented in Fig. 1. Its modifications were 
used in the construction of the following lenses: 12.7 mm //2.8, 
25 mm//2.7, 37.5 mm//3.5, and 50 mm//3.5. 

The excellent state of correction of these constructions may be 
illustrated by comparing the residual aberrations of the 12.7 mm//2.8 
Animar with the Rayleigh-Conrady tolerances (Table II). The 
following reservations should be noted with regard to this comparison. 
In deriving the formulas for tolerances, Conrady subjected the Ray- 
leigh criterion to certain interpretations, and he made certain assump- 
tions which are not necessarily realized in every design and under 


the actual conditions of focusing a lens. Thus the tolerance for 
marginal spherical was derived on the assumption that this aberration 
is preponderantly primary in a lens under consideration; this, how- 
ever, practically never occurs in well-corrected lenses in the range of 
speeds represented by the Animar series. The tolerance for zonal 
spherical was based on the assumption that the marginal spherical 
is reduced to zero, which is not a generally preferred state of correc- 
tion. The derivation of the tolerance for astigmatism and curvature 
of field involved the supposition that there is no vignetting and that 
the lens is focused on the image plane midway between the axial and 
the marginal focus. Since all these conditions are hardly ever satis- 
fied simultaneously, the Rayleigh-Conrady tolerances may be rigor- 





Marginal spherical 
Zonal spherical 

0.08 mm 
0.11 mm 

+0.01 mm 
-0.09 mm 


strict 0.06% 

Comatic patch: 

practical 0.25% 
strict 0.005mm 


practical f . 025 mm 

0.008 mm 


strict 0.01 mm 

practical f 0.14 mm 

T -0.05mm 

S 0.09mm 

* All the aberrations are for D spectral line. 

The oblique aberrations are for 12.5 field angle, 
t For extremely sharp definition. 

ously applied only in some special cases of extremely well-corrected 
systems. The other extreme is represented by the systems with re- 
sidual aberrations many times greater than the Rayleigh-Conrady 
tolerances. Then, no useful information can be obtained from refer- 
ring to these tolerances. When, however, a system is so well cor- 
rected that its residual aberrations approach the tolerance values, a 
juxtaposition of the residuals and the tolerances should serve well at 
least for illustrating the merits of the design, if not for its quantitative 

The 15 mm //3.5 Animar for the 16-mm frame is the Tessar type. 
Its correction is almost as perfect as that achieved in the 12.7 Animar 
although the angular coverage in this case is nearly two times greater. 
Due to the lower speed, the spherical aberration is even better cor- 
rected than in the 12.7 Animar, the zone being less than one third of 


the Rayleigh-Conrady tolerance for //3.5 lenses. The coma is well 
corrected within the entire field, and in the very margin of the field 
it is still near the Conrady practical tolerance. Both tangential and 
sagittal curvatures are well within his tolerance. 

It was particularly difficult in this case to meet the coma tolerance 
because it was undesirable to resort to the usual expedient of reducing 
the diameter of the peripheral image-forming pencil and sacrificing 
the corner illumination. It was believed that uniformity of illumina- 
tion should be considered a very important requirement particularly 
since the advent of color photography. Hence, an attempt was made 
to provide a relatively wide-angle lens with a corner-to-center ratio 
of illumination as high as could be obtained without accepting undue 
compromises with the image quality in the margin of the field. This 
effort resulted in a formula with the illumination ratio of 40% at the 
field angle of 22. The authors believe this is a very satisfactory 
result, considering the fact that for ordinary //3. 5 Tessar constructions 
of short focal lengths the ratio is usually below 30%, and that among 
the many //3. 5 and faster lenses (of focal lengths from J^ to 5 in.), 
either in our sample collection or reported in literature, only one 
//3.5 lens 14 with the relative illumination as high as 50% at 22 was 
found. It should be noted that the problem of illumination distribu- 
tion usually becomes more difficult for faster lenses. Still, even for 
//1. 9 and //1. 5 lenses in the Animar series a relative illumination 
greater than 40% was secured at the margin of the usable field. The 
ratio for the slower Animars of longer focal lengths is, of course, con- 
siderably higher. It increases with focal length, and approaches 
100% for the 100-mm Tele-Animar. 

Four elements are used in the //1. 9 Animars. The sequence of their 
powers is plus-minus-minus-plus, which arrangement has been suc- 
cessfully utilized in a number of other designs (type L-c of the Kings- 
lake classification 15 ). The limitation imposed by the small number of 
elements was a serious obstacle in attempts to obtain a high degree of 
correction for these formulas. Nevertheless, a design was finally 
produced whose oblique aberrations, referred to an average plane of 
focus, do not significantly exceed the practical Conrady tolerances 
for //1. 9 lenses. In order to eliminate the disturbing effects of a focus 
shift when the lens is stopped down (this is a quite common phenom- 
enon in lenses for general photographic purposes), the marginal 
spherical aberration was left over-corrected, while the zonal spherical 
was reduced to less than two times the strict Rayleigh-Conrady limit. 




In the authors' judgment, the over-all correction of these lenses is 
equal to or better than that of other similar designs, and their photo- 
graphic performance was found highly satisfactory in Bausch & 
Lomb laboratories and in independent tests. 

Four-element constructions have been frequently used for //1. 5 
lenses, and their deficiencies have not been too severe at least in 
amateur motion picture photography. It was thought, however, that 
at the present level of the art, and particularly in professional applica- 
tions, a considerably higher degree of correction is needed than seems 
to be obtainable with four-element constructions 'of //1. 5 speed. 
For this reason six elements were used in the //1. 5 Animars. The con- 
struction is of the Gaussian type, whose origin and evolution are well 
described in a paper by A. Murray. 16 The excellent potentialities 
of this type of construction have been exploited in a number of suc- 
cessful designs, as may be represented by the well-known Baltars, 
Biotars, and some types of Ektars. 





Marginal spherical 
Zonal spherical 

Comatic patch: 

0.03 mm 
strict 0.03% 
practical 0.25% 
strict 0.003 mm 
practical f . 025 mm 
strict . 003 mm 
practical f . 08 mm 

0.03 mm 

0.047 mm 

T -0.02mm 
S -0.04mm 

* All the aberrations are for D spectral line. 

The oblique aberrations are for 11.0 field angle, 
t For extremely sharp definition. 

The formula of the //1. 5 Animars is closely related to that of the 
//1. 6 Super Cinephor projection lenses for 16-mm film, which were 
introduced by A. Neumer to the preceding SMPE convention. 17 
Major design work was, however, required for obtaining a satisfactory 
state of photographic correction at the //1.5 speed. The correction 
of the basic formula is illustrated in Table III. 

The formula utilized in the design of the 75 mm and 100 mm //3.5 
Tele-Animars may be classified as a triplet construction with the 
front element split into two elements, the second of which is a menis- 




cus. A prototype of this construction may be traced back to a 
formula produced in 1925 by L. Bertele. 18 This prototype was later 
subjected by a number of designers to extensive modifications some of 
which involved the introduction of additional elements and their 
compounding. From these modifications evolved the Sonnar lenses 
whose excellent characteristics are generally recognized. It seems, 
however, that the highly favorable possibilities offered by the basic 
four-element construction have been neglected for some time, particu- 
larly in this country, although the construction has been commercially 
utilized by the British. Only recently has a revival of interest been 
revealed by some designers. 19 

The triplet construction with the split front element is highly 
favorable to a satisfactory correction of spherical aberration at fast 
speeds, for an excellent correction of coma, and apparently for a 

Fig. 2. 100-mm Tele-Animar //3.5 in focusing mount. 

better correction of astigmatism and of curvature of field than is 
obtainable with other fast constructions utilizing only four elements 
(the Petzval and related types). However, the curvature of field still 
remains the basic limitation, so that the coverage can hardly be more 
than about 15 half-angle if strict criteria are used. This coverage is 
approximately represented by the minimum focal length of 10 mm 
for 8-mm frame, of 25 mm for 16-mm frame, and of 50 mm for 35-mm 

The excellent characteristics and the apparent limitations of the 
four-element construction were thoroughly analyzed in the course of 
the Tele-Animar design. On the basis of these studies, the conclusion 
was reached that, as compared with a simple triplet, the split triplet 
construction offers definite advantages particularly for covering the 
16-mm frame with lenses of focal lengths longer than 50 mm. Conse- 


quently, a four-element formula of the type just described was de- 
signed for the 75 mm and 100 mm Tele-Animars, illustrated in Fig. 
2. The speed of these lenses was limited to//3.5, although it easily 
could be made higher. The reason for this conservatism was that the 
almost ideal correction of the formula at //3.5 made the authors 
reluctant to accept even minor compromises just to gain in speed 
designation, especially considering the fact that a high speed is rarely 
of much importance in "telephoto" use. The speed of the 50 mm 
and 37.5 mm Animars was kept at //3.5 for the same reason. The 
correction of the Tele-Animar formula is summarized in Table IV. 





Marginal spherical 
Zonal spherical 

Comatic patch: 
Curvature : 

0.17 mm 
strict 0.04% 
practical 0.25% 
strict 0.006 mm 
practical f . 025 mm 
strict 0.01 mm 
practical f 0.18mm 

+0.07 mm 
-0.07 mm 

0.013 mm 

T +0.03 mm 
S -0.03mm 

* All the aberrations are for D spectral line. 

The oblique aberrations are for 3.5 field angle, 
f For extremely sharp definition. 

In the preceding discussions pertaining to the correction of the 
Animar series, nothing was said about its color correction. This is 
because a satisfactory color correction is regarded as an obvious pre- 
requisite of any modern lens intended for motion picture photography. 
While this requirement usually is met without serious difficulties, no 
magic procedures can be derived for obtaining a "superior" color 
correction of fast lenses covering a substantial field, because a limita- 
tion is basically established by the now available selection of optical 
glasses. Therefore, a cursory remark should suffice: that the sec- 
ondary spectrum for every lens in this series was reduced to the gen- 
erally anticipated and accepted limits, and that the oblique color was 
not permitted to exceed a few microns even with the leiises of longer 
focal lengths. 

Finally, it should be noted that, in appraising the performance of a 
lens, the user naturally cannot disregard its behavior at other stops 


than the maximum, unless his intentions are to deal only with some 
special situations requiring all the available speed of the lens. In this 
connection, observations are frequently made to the effect that, be- 
cause certain compromises are unavoidable in fast lenses of any de- 
sign, a fast lens stopped down should be necessarily somewhat inferior 
to a lens specially designed for use at that lower stop. There are no 
inherent factors that would make such a situation generally valid. 
Indeed, with a favorable balance of its residual aberrations, a lens 
should improve when it is stopped down, and at the smaller stops it 
may be as good or even better than any other specially designed slower 
lens. This situation actually prevails in the Animar series. As 
any one of these lenses is stopped down, its residual aberrations do 
rapidly diminish, and at a stop in the neighborhood of //5.6 practically 
all of them are brought substantially under the corresponding strict 
Rayleigh-Conrady tolerances. This advantageous characteristic of 
the Animars follows almost self-evidently from the fact (illustrated 
in the preceding tables) that their residual aberrations even at the 
maximum apertures are either under or comfortably near the respec- 
tive tolerances, and it may be further substantiated by the following 
numerical data. The Rayleigh-Conrady tolerance for the marginal 
spherical at //5.6 is 0.31 mm; the tabulated data indicate that this 
tolerance is much larger than the residual spherical even of the wide- 
open Animars; when they are stopped down to //5.6, the residual 
becomes entirely negligible in comparison with the corresponding 
strict tolerance. The Rayleigh-Conrady strict tolerance for coma at 
//5.6 is 0.01 mm; for the 12.7 mm//2.8 formula at//5.6 the trigono- 
metrically computed comatic patch is 0.005 mm (less than the Airy 
disk, whose diameter at //5.6 is 0.008 mm) ; the patch for 15 mm 
//1.5 at//5.6 is 0.014 mm and for 100 mm//3.5 at //5.6 it is 0.009 
mm. The strict tolerance for curvature of field at //5.6 is 0.04 mm 
and the practical (for extremely sharp definition) is 0.28 mm; the 
actual residual curvatures of the Animars (see the tables) are within 
these limits even with the lenses wide open. 

The authors do not claim that some special procedures were used 
for obtaining this favorable state of affairs. The gratifying results 
should be attributed to perseverance and luck, which are indispen- 
sable, though not omnipotent, ingredients of any successful optical 



Accomplishments in any large organization seldom can be attributed 
to only one or two individuals. Credit is gratefully given: to 
Misses L. M. Hudson, B. N. Marble, and L. B. Frey for their exten- 
sive contributions in the optical development of the Animar formulas; 
to Mr. C. DeGrave, who was primarily responsible for the mechanical 
design; and to Mrs. M. H. Tarplee for her patient measurements of 
the experimental samples. 


( 1 ) W. B. Rayton, "A new series of camera lenses for 16-mm cinematography," 
Jour. SMPE, vol. 48, pp. 211-216; March, 1947. 

(2) G. A. H. Kellner, "Projecting lamp," U.S. Patent No. 969,785; September, 

(3) G. Gamow, "Any physics tomorrow?", Physics Today, vol. 2, pp. 16-21; 
January, 1949. 

(4) W. Taylor and H. W. Lee, "The development of photographic lens," 
Proc. Phys. Soc., London, vol. 47, part 3, pp. 502-518; May, 1935. 

(5) K. Pestrecov, "Resolving power of photographic lenses," PSA J., vol. 
13, pp. 155-158; March, 1947. 

(6) K. Pestrecov, "Resolving power of photographic lenses," Photogram. Eng., 
vol. 13, pp. 64-85; March, 1947. 

(7) F. H. Perrin and H. 0. Hoadley, "Photographic sharpness and resolving 
power," J. Opt. Soc. Amer., vol. 38, pp. 1040-1053; December, 1948. 

(8) D. G. Hawkins and E. H. Linfoot, "An improved type of Schmidt camera," 
Mon. Not. R. Astr. Soc., vol. 105, pp. 334-344; December, 1945. 

(9) M. Herzberger, "Light distribution in the optical image," J. Opt. Soc. 
Amer., vol. 37, pp. 485-493; June, 1947. 

(10) A. Marechal, "Etude des effects combine's de la diffraction et des aberra- 
tions geometriques sur Timage d'un point lumineux," I, II, III, Revue d'Optique, 
vol. 26 and 27, pp. 257-277; 74-92; 270-287; September, 1947; February and 
May, 1948. 

(11) O. H. Schade, "Electro-optical characteristics of television systems," Parts 
I, II, III, IV, RCA Review, vol. 9, pp. 5-37; 245-286; 490-530; 653-686; March, 
June, September, and December, 1948. 

(12) A. E. Conrady, Applied Optics and Optical Design, Oxford Univ. Press, 
London, 1929; pp. 137, 393, 433. 

(13) K. Pestrecov, "Three element objective lens," U.S. Patent No. 2,453,260; 
November, 1948. 

(14) F. Benford, "Illumination in the focal plane," /. Opt. Soc. Amer., vol. 31, 
pp. 362-368; May, 1941. 

(15) R. Kingslake, "A classification of photographic lens types," /. Opt. Soc. 
Amer., vol. 36, pp. 251-255; May, 1946. 


(16) A. E. Murray, "The Baltar series of lenses," Intern. Photo., vol. 19, pp. 
12-14; June, 1947. 

(17) A. E. Neumer, "New series of lenses for professional 16-mm projection," 
Jour. SMPE, vol. 52, pp. 501-508; May, 1949. 

(18) Ernemann-Werke, "Photographic lens," British Patent No. 237,212; 
October, 1925. 

(19) P. E. Creighton, "Lenses for cinematography," U.S. Patent No. 2,432,387; 
December, 1947. 


MR. HERBERT LOWEN: Can you tell us whether the field curvatures were com- 
puted along the principal ray or for the full fan? 

DR. PESTRECOV: As usual, correction of curvature as shown in the slides was 
computed along the principal ray; but, of course, the residual coma contributes 
to the effective curvature. We try to over-correct the fan to compensate for any 
under-corrected residual curvature along the principal ray or vice versa. If we 
have over-correction of the principal ray, we try to under-correct the fan at that 
particular angle to compensate for that, with the aim of providing as flat a field 
as possible with the wide open beam at the particular point in the film. 

MR. LOWEN: And the tangential difference was for the corner? 


MR. LOWEN: And what was the maximum astigmatic difference? 

DR. PESTRECOV: In most of these lenses, because the angles are small, the corner 
really represents the maximum difference. 

Color Cinematography 
In the Mines 



Summary Problems involved in 16-mm color cinematography in the 
mines are discussed. Details of power supply, distribution, and voltage 
control are explained. Various photographic techniques necessary to the 
obtaining of exposure, color quality, and modeling under adverse conditions 
are illustrated. 

COLOR CINEMATOGRAPHY in the mines is a challenge to the imagina- 
tion as well as to the physical endurance of those who undertake 
it. Of ten assorted mines vi'sited by the author, over a traveled dis- 
tance of 11,000 miles, a salt mine lent itself most favorably to photo- 
graphic requirements. Nevertheless, it, too, offered those basic ob- 
stacles peculiar to other similar ventures underground. 

Entering a mine with two tons of photographic equipment was a 
laborious and tedious job. Skip hoists were small, limiting a load to 
only a few items. To lower eight 5,000-watt Senior spotlights with 
standards required four trips in many instances. In addition there 
were six 2,000-watt Junior spotlights, four Broadside Doubles 1 flood- 
lights, hundreds of feet of cable, transformers, and the usual array of 
carrying cases. 

After the equipment was lowered to the working level of the mine on 
the skip hoist it was loaded on a train of cars for the trip toward the 
scene of operations, usually a mile or more from the shaft. Arriving at 
a transfer point it was then reloaded to a number of shuttle cars and 
drawn to the "face" of actual mining operations. Thus, many hours 
were spent moving about within the mines. 


In the salt mine the "roof" rose from ten to twenty feet permitting 
lighting from parallels constructed of available workbenches and 
boxes. The uneven floor of a room was leveled where each lamp was to 
be placed and the casters were set in channels cut with a pick. 

Since the source of power was usually within a hundred feet of the 

PRESENTED: October 11, 1949, at the SMPE Convention in Hollywood. 


200 M. CHARLES LINKO February 

shooting area at the mine face, the anticipated voltage drop through 
loss in cables, in addition to the load drawn by the mining 
machinery, suggested the use of the Color Photographic Type 
115- volt, 3350 K (degrees Kelvin) lamps with Kodachrome Commer- 
cial Film. Because this emulsion is balanced for 3200 K it required 
only further reducing the voltage, and consequently the color tem- 
perature, in order to satisfy its color balance characteristics. 

In this mine, supplied with 460-volt, three-phase alternating cur- 
rent, it was necessary to use two 15-kva, 4-to-l, double primary and 
secondary type transformers connected open delta. This permitted 
connection of two 100-ft three-wire cables with two six-hole plugging 
boxes to the secondary terminals of each transformer. These were 
placed in close proximity to the lamps. Fortunately, telephone com- 
munication with the power plant and excellent co-operation from the 
mine officials made it possible to obtain the desired voltage for any 
setup, thereby solving the major problem in this mine. 

Color temperature control was accomplished by referring to the 
lamp manufacturer's color temperature factor curve for the proper 
voltage required to operate the lamps which indicates a color tempera- 
ture change of approximately 10 K/volt. Accordingly, the 102 volts 
necessary to yield 3219 K was adhered to as closely as circumstances 
permitted. A rise to 108 volts or 3276 K was tolerated. On close set- 
ups numerous lamps were allowed to remain burning outside the area 
being photographed in order to prevent a voltage rise beyond the de- 
sired level. The voltage readings were made at the lamp base. The 
consequent loss of light intensity resulting from this manner of volt- 
age control was overcome by the use of the Maurer Professional Cam- 
era with its 235 shutter opening. The key light was set at 500 ft-c at 
//2.8 as compared to 900 ft-c recommended by the film manufacturer. 2 
Working with a lower level obviated the necessity for intense spotting 
of the light beams, or the alternative of moving in with the lamps, 
either of which would have limited many long shots. The effect was 
that of even over-all illumination without crowding the narrow rooms 
and passages with a battery of floodlights. 

Kicker lights or crosslights were added to enhance the scenes but 
only in rare instances was backlighting possible. To obtain the illu- 
sion of depth, a slightly more intense beam, properly directed and dif- 
fused with black scrim, was played on the background from the side, 
thus preserving the rugged texture with slight shadow detail. It was 
not uncommon to conceal a lamp in the set behind the very machine 


being photographed, often lowering the lighting unit into a pit ex- 
cavated for the purpose. 

A common difficulty was that of lighting a long, narrow tunnel 
where the far wall was to be seen in the picture, yet much of the action 
was to take place within the passage. This action often included a 
huge piece of mining machinery passing by and taking up all of the 
passageway but a few feet on either side. In order to obtain sufficient 
light for exposure, niches were made in the ribs and the lamps con- 
cealed within them ; light beams were reflected off the roof and the dis- 
tant ribs; a lamp was tied to the back of the moving machine; front 
lights were crisscrossed and flooded, or gradually diffused and reduced 
in intensity with black scrim as the machine approached the camera 
every artifice was put to use. The extreme range of light level re- 
quired for the scene before the action had taken place, as opposed to 
that required during the enactment, taxed all available facilities. 
Nothing short of remote-controlled shutters on the lamps would have 
served to better the condition. 

Moisture combined with salt particles in the air to form hydrochlo- 
ric acid, requiring that the camera and accessories be dubbed with lan- 
olin to ward off corrosion. Lenses were constantly being dusted. The 
emulsion chosen required no color correction filter, thereby eliminating 
the troublesome annoyance of caring for dusty or fogged filter sur- 

At the end of each shooting day all equipment had to be broken 
down and moved to a passage known to be idle overnight because work 
crews progressively blasted in each section. Frequently it was nec- 
essary to load it on cars, have it hauled to a safe passageway, only to 
haul it back to the same place the following morning. 

Upon completion of shooting in this mine, it was noted that the ac- 
tion of the salt had extensively attacked the metal standards, plugging 
boxes, and the like, necessitating a thorough cleansing before leaving 
the location. 

The footage made in the salt mine, aided by the soft reflective na- 
ture of the mineral, was of "high key" value. The slight blue-gray 
tint in the rock was characteristically reproduced in favor of the blue 
tending toward the green in the shadows. The gray values seldom oc- 
curred as a true mixture of black and white. The orange-lacquered 
mining equipment added to the beauty of the settings and was shown 
to excellent advantage in the wide corridors. 





In the potash mines the rooms, ranging from eight to twenty feet in 
width, were less spacious than in the salt mines and the roofs, about 
ten feet high, were lower. The mineral was salmon colored, with 
strata of salt producing what might be a child's conception of "Candy- 
land." Like salt, potash, being almost opalescent, does not reflect 

Fig. 1. The more spacious salmon-colored potash mine, where illumination 
difficulties arose from surges caused by the operation of heavy mining machinery 
such as the shuttle-car, a portion of which can be seen in the right background 
(Photo by Tom Toia). 

The opening scenes were to be filmed in rooms adjacent to those 
being mined. The heavy loading machines drained the power sup- 
ply, causing a fluctuation in the illumination. The only recourse was 
to light the set with this wavering illumination, instruct the personnel 
in the action desired, and when all was in readiness for the camera to 
roll, to signal the operator in the adjoining room to stop his machine. 
A cursory survey of the lighting was made before the take, and in this 


manner the shooting progressed. The mining machine being photo- 
graphed was equally offensive from the power usage standpoint, how- 
ever, and only in the final cutting does the illumination remain con- 
stant in effect. The initial dimming was of necessity tolerated because 
the cost of running a separate line into the mine was prohibitive. 
Resting on the surface was a 20-kva generator used for lighting the 
5,000-watt Seniors when used as boosters on exterior lighting, but 
which was of no value in this predicament because it could not be 
transported close enough to the scene of action. 

Even more distressing was a bad "take" because the area became 
partially "mined out" in the process of the shooting and to maintain 
continuity the fragmentary ore had to be shoveled back into place for 
the retake. It may be added that a gunshot would pass unheard 
among the thunderous sounds of a mine in operation. Conversation, 
for the most part, was carried on by one's newly developed sign lan- 

A specific type of cutting machine was encountered in a potash 
mine where the rooms were extremely narrow, scarcely wide enough to 
permit passage on either side of the mechanical monster. There was 
but one direction in which to shoot and that was toward the mine face. 
Since the cutter's bar also worked its way forward, there was no space 
left for lamp placement. Closer observation of the area revealed that 
advancing drilling crews had left some holes in the ribs just below the 
roof. These served to accept some quickly improvised dowels and 
wire hooks from which 5,000-watt Seniors were suspended. The light 
beams were directed at the face allowing some light to spill over on the 
ribs. Two Seniors placed further back and to one side were diffused 
with black scrim to an intensity of 300 ft-c and directed at the work- 
men in the room. The camera was mounted on two boards elevated 
on sawhorses just a few inches above the incoming machine. The 
forced perspective of the 17-mm lens produced a striking elongated 
mine passage view as the cutting machine emerged from beneath the 
camera. Thus, a hopelessly restricted area was instrumental in pro- 
viding an effective coverage which might have been otherwise 

It was becoming increasingly appparent that lenses of the shorter 
focal lengths were the order of the day. A 20-mm lens was needed to 
supplement the rather extreme 17-mm wide angle that had been ade- 
quate in the more spacious mine. Such a lens, on order for many 
months, had been delivered to the distributor in the East. The cam- 

204 M. CHARLES LINKO February 

era was flown to the distributor while exterior scenes were made with a 

The color obtained in the potash mines was indeed faithful to the 
original. The emulsion used favored those subtle tints within the 
warm region of the spectrum. Shadows were void of the green veiling 
that prevailed in the salt mine footage. 


The intense humidity of the iron mines, located 300 ft below sea- 
level, and the ever-present staining of red iron oxide dust caused much 
discomfort and grief. The moist particles became imbedded in the 
crackle finish of the camera parts, necessitating a daily scrubbing 
with soap and water, all other solvents having failed. Personnel ap- 
peared to be wearing a deep shade of panchromatic make-up which 
luckily was photographically of material aid to their dark skins. 

The dimensions of the rooms were much like those of the first pot- 
ash mines. Lamp placement was more difficult here because the ore 
was less thoroughly pulverized. Huge piles of the heavy muck were 
scaled and leveled before the lamps could be set up. 

Of greater consequence was the inadequate power supply. The 
mean average output was approximately 98 volts. Surges from vari- 
ous motors in use caused a drop of from 10 to 15 volts which was made 
doubly bad by the low reflection coefficient of the dull red surround- 
ings. The scenes were lit in accordance with the intensity required al- 
lowing the light to go warm on the premise that the predominant color 
would serve as a camouflage. 

Ninety-eight volts, or 3169 K, represented a departure of lesser con- 
cern at this time, for the new low voltage level drastically reduced the 
light intensity as well. 

In order to obtain the 500 ft-c required for the flesh tones, the light 
beams from two Seniors placed in tandem were overlapped. Back- 
grounds were first evenly flooded, then the projections and recesses 
were crosslighted to accentuate the differences in the various planes. 
Takes were made until one was obtained where the power remained 
fairly constant, the ponderous ore being replaced after each take. 
The gray of the steaming atmosphere was increased by the exhaust 
from the air-powered drills and upon first inspection appeared to be a 
photographic impossibility; however, the use of a minimum of cross- 
lights, essential to impart a feeling of depth, eliminated the fog. The 
effects of the lower voltage were indiscernible in the print. The moist 


red rock is reproduced in vivid tones of monochrome and, strangely 
enough, some of the deeper shadows are veiled with a transparent 
film of blue. 


The zinc mines presented the same conditions encountered in the 
iron mines but with one addition: water. Water rushed in torrents 
along the tracks; it formed puddles throughout the mine ; it dripped 
off the roof and made a sea of mud of the working areas except for the 
very face of operations. The roof was so low here that one had to 
walk with his head tilted sidewise ; then suddenly the clearance rose to 
a height of 75ft. 

The power supply dropped to an all-time low, the highest reading 
observed being 96 volts. Since this mine was supplied with 250-volt 
direct current, the power was tapped from the nearest trolley line. 
One lead, fitted with a fusable miner's nips (a rod shaped in the form 
of a question mark with an insulated handle at its base) was hung 
on the trolley line, the other being clamped to the rail. The Seniors 
were connected in series of two units each, making a set of four series 
units, or eight lamps, parallel across the line and the whole was con- 
trolled by a master switch. 

Since the rock being mined was largely limestone with thin serrated 
strata of yellow and silver zinc, the dull gray surroundings required a 
great deal of light. "Effect lighting" was resorted to in the long shots 
resulting in a more realistic representation perhaps, but one that de- 
parted somewhat from the style conventionally used in industrial 
filming. Nevertheless, many takes were made before the power re- 
mained sufficiently constant throughout any single operation. 

Some of the workings were photographed at night when only a small 
portion of the mine was functioning. Then there was an overabun- 
dance of power. The use of the variable mine car resistor at full capac- 
ity, with three Seniors burning off the set for voltage control, pro- 
vided 3240 K. 

The control of color temperature by voltage regulation was pre- 
ferred to the use of color compensating filters because no method was 
available at the time for accurately evaluating filters in terms of de- 
grees Kelvin. The feasibility of constant use of color compensating 
filters in the mines is questionable but, as an adjunct to partial volt- 
age reduction, no doubt the aid of the proper filter was direly 
needed. The results in this mine attested to this fact because the cool 
gray surroundings reproduced with a steel-blue cast. 




The natural ventilation in a limestone mine with its entrance at 
ground level permitted driving the truck and generator to the very 
face of the workings. Thus, after 10,000 miles of travel, the genera- 
tor finally was put to use within the mine proper. Without this ideal 
generator location, shooting would have been impossible because the 

Fig. 2. The main haulage-way in the limestone mine, where the natural venti- 
lation permitted the use of a generator driven by an internal combustion engine; 
note the Seniors in the center middle distance placed on the rocky ledge to the 
left of the string of cars (Photo by Beryl Hawkins). 

power available from the mine equipment was little above that re- 
quired to operate the mining machinery. The generator supplied 
power for six of the Seniors, the mine circuit taking the other two and 
some of the smaller units. With virtually complete control of illu- 
mination, shooting day and night, the crew finished in less than half 
the time consumed in other mines. The results here were consistent 
with the more favorable conditions. 





The three coal mines visited were supplied with direct current where 
in all cases it was necessary to reduce the voltage. With all lamps 
burning and the resistor in the circuit, in one instance a length of iron 
pipe tied into the line served to lower the output. 

Some of the Color Photographic globes were replaced with the 3200 
K Motion Picture Type and their beams were blended. The 2,000- 
watt Junior spotlamps were so used to impart a warm tone to the 
black mineral. 

Fig. 3. Shooting a scene of a cutting machine at the "face" of operations. 
Photo shows nearly the full width of a typical room in a coal mine where the heat 
of the lamps dried the surfaces, causing slabs of coal to drop all about, and where 
the dust obliterated every scene, often necessitating leaving the rooms until the 
atmosphere cooled and cleared (Photo by Beryl Hawkins). 

The major difficulty in a coal mine is that of balancing the flesh 
tones with the dark surroundings. When the coal surfaces appeared 
without the beneficial, reflective layer of rock dust, 1,600 ft-c were re- 
quired on the coal surface to accomplish this balance. Moreover, the 
surface was deliberately marred to create a texture of minute colorful 


The timber supports used in the coal mines add to the lighting dif- 
ficulties because they, too, are light in color, although they served well 
to conceal many lighting units. 

Another difficulty was that of keeping the participants within pre- 
determined boundaries. The miners would inadvertently walk into a 
"hot" beam of light intended for the illumination of a mound of coal. 
The problems experienced in other mines were amplified here because 
of the absence of bright surfaces from which to reflect the light rays. 
More dust prevails in a coal mine. In addition, all mine surfaces, 
save the floor, are covered with rock dust and this fine abrasive is in 
constant circulation. The intense heat of the lamps sometimes dried 
the roof, causing slabs to drop all about. This was more frequent in a 
low mine where the roof was a mere forty inches above the floor. At 
such a time it was necessary to leave the area to allow the surfaces to 
cool. Moving about on one's haunches for a number of hours added 
physical pain to the other difficulties. 

The long scale gradation of Kodachrome Commercial Film was un- 
doubtedly most appreciated upon viewing the footage made in the 
coal mines. The full brightness range was covered with magnificent 
detail in both extremes. There is a mild coolness about this footage 
much of which was photographed at 3240 K. Black dust particles in 
the air and indeed the cold nature of the surroundings were contribut- 
ing factors. Backlighting on a stream of coal imparted a rust-red hue 
to the mineral, otherwise it photographed as a true black with vari- 
colored catch lights. 


Obviously the major problem in mine cinematography is the some- 
what indefinite power availability. Except in some naturally venti- 
lated mines, generators, driven by internal combustion engines are 
prohibited. A voltage stabilizer would have been an asset, for very 
little photography may be done at night when the power is in lesser de- 
mand. Nevertheless, cinematography in the mines is fascinating for 
there is much color there. True, it does present seemingly insur- 
mountable obstacles, but, when these are overcome, one has the satis- 
faction of having won over tremendous odds. 


(1) R. G. Linderman, C. W. Handley, and A. Rodgers, "Illumination in motion 
picture production," Jour. SMPE, vol. 40, p. 333; June, 1943. 

(2) Eastman Kodak Company pamphlet, Kodachrome Commercial Film Type 

Television Test Film 

WITH THE EXPANSION of commercial television operation at the 
close of the war there arose a need for detailed information on 
the production and use of films in television broadcasting. The 
Television Committee undertook to meet this need by approaching 
the problem from two directions. 

First, they compiled information on what was believed to be good 
practice in the production of television films; and second, they 
provided the television broadcasters with a means of adjusting their 
film camera chains to best reproduce these films. 

The first step in this combined project was completed in February, 
1949, when the report "Films in Television" was published and dis- 
tributed throughout both the television and motion picture indus- 

The second phase of the work was completed in September, 1949, 
when a subcommittee under the chairmanship of Dr. R. L. Garman 
approved the television test film for release. Every effort was made 
by the subcommittee to incorporate in this film, which is now availa- 
ble to broadcast stations and equipment designers, the ideas and 
suggestions advanced by the television industry. 

It is hoped that the adoption of the limits and procedures defined 
will aid the television broadcasters in raising the general quality 
level of film program transmission. 


The test film is designed to indicate the condition of operation of 
those portions of the television film reproduction system which de- 
pend upon the relation between the film projector and the television 

Use of the test film on a routine operational basis is recommended 
since it will indicate errors of adjustment and equipment malfunc- 
tion before they might otherwise be detected. 

To facilitate making extended service adjustments, or to provide 
a suitable subject for the initial setup and adjustment of a film 
channel, there are available separate lengths of the alignment, low 
frequency, storage, and transfer characteristic sections. These 
sections may be cut into appropriate lengths, made into loops, and 
run continuously as the need arises. 




Figure 1 

Figure 2 


The film is not intended to be a laboratory instrument, although 
it may be useful in product design and test. 


Seven test sections and a selection of scenes comprise the complete 
film which is available in either 16- or 35-mm widths. The test sec- 
tion is a series of geometrical patterns intended to present informa- 
tion on the factors most likely to be degraded in television film re- 
production. Each chart selects some particular failing of the aver- 
age system and produces a signal intended to exaggerate and thus 
clearly define any deviation from normal operation. Perfect re- 
production of all the charts is to be desired, but some degradation 
of each is to be expected. Experience will show the magnitude of 
these effects which may be considered normal for any particular 

Scenes representative of many types of pictures encountered in 
television films are included in the reel as a final qualitative test of 
over-all results. 

Sec. 1. Alignment (See Fig. 1) 

This pattern defines the portion of the projected film frame which 
is to be reproduced by the television system and permits accurate 
alignment of the motion picture projector with the television 
camera. Eight arrow points have been positioned to touch the 
edges of the picture area to be scanned. This area is smaller than 
that of the whole frame. One and a half per cent of the projected 
aperture is cut at top and bottom of the frame to allow for small drifts 
in scanning and centering. The horizontal dimension is chosen to 
provide a standard four-to-three aspect ratio with the established 
height. All of the frame area beyond these limits has been striped 
with a " barber-pole" effect. This striping must not appear in the 
television picture. 

It should be noted that the striped area is wider on the sides of 
the frame than on the top and bottom. This results from the fact 
that the standard projection aperture does not have a four-to-three 
ratio but is wider by some 3%. See the American Standards for 
Picture Projection Apertures, Z22.58-1947 and Z22.8-1950. 

Each vertical arrow head is 4% of the picture height and each 
horizontal arrow head is 4% of the picture width. Similarly, the 
arrow shanks are 6% of the picture height and width respectively. 




Figure 3 

Figure 4 


These dimensions permit rough estimates of the magnitude of scan- 
ning irregularities or misalignment through visual comparison of the 
effects in question with the size of the arrows. Specific values for 
misalignment obtained in this manner can be logged easily for future 
reference as part of a quality control program. 

A rectangle formed by the lines connecting the arrow shanks 
encloses 80% of the active picture area. Investigation indicates 
that this area is reasonably well reproduced on most home receivers, 
even in the presence of scanning drift, inaccurate adjustment, and 
abnormal masking. No standard is implied but a general memory 
of this area may be useful in the preparation of film carrying impor- 
tant information. 

Sec. 2. Low-Frequency Response (See Fig. 2) 

This test is made in two parts, each consisting of a half -black-half- 
white frame, with the dividing line horizontal. The first section has 
the black portion at the top of the frame and the second is black at 
the bottom. These charts produce 60-cycle square wave signals. 
When viewed on the wave-form monitor set for field rate deflection, 
the signals should appear reasonably square. Serious tilting or 
bowing indicates incorrect low-frequency phase and amplitude re- 
sponse. When the system has been set for reproducing the first 
chart, the change to the second chart should not necessitate large 
shading changes. 

The chart which is black at the bottom also permits a check on the 
amount of flare encountered in Iconoscope operation. Rim lights 
and beam current should be reset if the flare is excessive. 

Sec. 3. Medium-Frequency Response (See Fig. 3) 

The response of the television system to medium-frequency sig- 
nals is of importance to picture quality. In this test, horizontal 
bars are used, first as black on white and then reversed. The bars 
have lengths equal in time of scanning beam travel to 2, 5, 12^, 
and 32 microseconds. These correspond to half-wave pulses cover- 
ing an approximate fundamental frequency range from 15 to 250 
kilocycles. Correct medium-frequency phase and amplitude re- 
sponse will be indicated by leading and trailing edges of the bars 
having no long, false gray tones. If, following the trailing edge of 
a bar, a streak appears having a tone similar to that of the bar 
(white after white, black after black) then it is reasonable to assume 




Figure 5 

Figure 6 


that the amplitude of the frequency represented by that bar is too 
great, or that its relative phase is incorrect. If the opposite occurs, 
as a white streak after a black bar, the fundamental frequency is 
too low in amplitude, and its relative phase is in error. 

Sharp transient effects immediately following all bars are an indi- 
cation of excessive high-frequency response. This condition will 
usually be clearly indicated in the test for resolution later in the film. 

If very long streaking occurs in which the spurious signals are 
seen on the left side of the bars, as well as on the right, an investiga- 
tion of the low-frequency response of the system should be made. 
Under these conditions close examination of the previous charts 
should reveal errors of wave form. 

It is rarely possible to obtain perfect streaking-free reproductions 
of both the black-on-white and the white-on-black charts with one 
setting of the controls. This may be due, with Iconoscope opera- 
tion, to the effects of wall sensitivity. A change in bias-light is 
usually required to compensate the charts exactly since the two 
charts do not have the same average transmission. The settings 
which produce very small streaking equally on both charts are usu- 
ally preferred. 

Sec. 4. Storage (See Fig. 4) 

Film pickup systems which utilize short pulses of light must store 
the charge produced by the pulse long enough to permit the charge 
image to be scanned. Since the beam starts the scanning process at 
the top of the picture, the storage time required is maximum at the 
bottom of the picture. Some pickup tubes will suffer from leakage 
to the extent that the charge image may be seriously reduced in 
amplitude by the time the beam reaches the bottom of the picture. 

The chart which checks this characteristic is made up of vertical 
black and white stripes on a gray background. When viewed on 
the wave-form monitor (set at field rate) this pattern will produce 
three lines representing white, gray, and black. Shading should be 
set to hold the gray line parallel with the blanking axis. If the 
white and black lines then tend to converge, the pickup tube does 
not have perfect storage. Perfect results are indicated when all 
traces are parallel. If the black-to-white amplitude at the bottom 
of the picture is divided by that at the top of the picture, the tube's 
storage factor is obtained. This is usually expressed in percentage. 




Figure 7 

Figure 8 


Sec. 5. Transfer Characteristics (See Fig. 5) 

The ability of a television system to reproduce shades of gray is 
indicated in this section through the use of step-density areas. The 
first chart consists of a white area and a black area that serve as 
limit references, along with a centrally placed window in which the 
density steps appear. 

The neutral gray background of this chart should be shaded flat 
and contrast and brightness settings adjusted to give normal wave- 
form monitor amplitudes from the reference areas mentioned. The 
wave-form monitor should be set for line frequency. Once adjusted, 
all settings should remain untouched during the remainder of the 
period. In the center window, a total of seven density tabs will 
appear labeled A through G. These will be seen in groups of three, 
as ABC, BCD, CDE, etc. This permits all steps to be read on the 
same portion of the mosaic and independently of shading and black 
spot. Each step should be visually compared with the adjacent 
steps, both in the picture and on the wave-form monitor, and each 
should be clearly defined. Saturation effects will be seen as a 
cramping together of adjacent steps. Experience as to the appear- 
ance of the tabs will establish a norm from which variations can be 

The final chart in this section consists of two step density tablets 
showing all seven steps together. The direction of progression of 
the second tablet is opposite to the first. These permit rapid over- 
all check. 

The effective transfer characteristic of a film pickup system is a 
function of both film density and projected illumination. This test 
film has a range considered to represent that normally encountered 
in practice. If significant compression occurs, projector brightness 
should be checked. Other factors, including beam current, bias- 
light, and clipper adjustments should be tested with a stationary 

Sec. 6. Automatic Brightness Control (See Fig. 6) 

This test indicates the ability of the television system to follow 
changes in average illumination of a series of scenes. It consists of 
a white disk centered in a black frame which enlarges slowly to fill 
the whole frame. As the white portion becomes larger, the bright- 
ness control should hold the black level constant. On the wave- 


form monitor, the black signals should remain fixed in position rela- 
tive to the blanking level. The first brightness changes on the film 
are both slow and even, so that systems with slow-acting control 
should be able to follow them accurately. 

The second portion of the tests consists of sudden changes in 
white disk size from the smallest to one-third frame area and then 
to two thirds of the frame area. Experience will show how much 
error in black level setting results in these cases on a transient basis. 

Sec. 7. Resolution (See Fig. 7) 

Each of the five charts in this section is carefully calibrated to 
indicate the over-all system response at the number of lines printed 
in its center. Starting at 200 lines, the charts change at five-second 
intervals until the 600-line pattern appears. Each chart permits 
reading the response at six points within the frame. Care should 
be taken to note the response at the edges, as well as the center, of the 

Under abnormal conditions of "lateral leakage," resolution of a 
stored-charge picture degrades with time. This condition can be 
evaluated by noting the relative top and bottom resolution. If 
there is significant difference between the two, the system should 
be checked with a continuously illuminated slide. If the slide test 
shows the same resolution at both top and bottom and the film test 
does not, the pickup tube may be at fault. 

The above presupposes that the projector has been properly tested 
for its inherent resolution with the visual test films available for that 

Sec. 8. Typical Scenes (See Fig. 8) 

To provide a qualitative check on the over-all results to be ex- 
pected from good film, several scenes taken from material used 
specifically for television are included in the test reel. Utilization 
of this section will depend upon the operator's experience in judging 
acceptability and upon his memory of "how they looked before." 

Recommendations for 

16-Mm and 8-Mm Sprocket Design 

Two major points affecting film life in 16-mm and 8-mm projectors are the shape of the film 
path entering and leaving the sprocket and the shape of the sprocket tooth. The following recom- 
mendations, approved by the Society, have been developed to assist the design engineer in specify- 
ing these, as well as other dimensions for particular applications. 


T A MEETING held on December 16, 1948, 
the Committee on Standards approved 
a recommendation made by its subcommittee 
on sprockets to rescind the existing standards 
on 16-mm and 8-mm sprockets because they 
were not flexible enough to take care of such 
variables as the shape of the path of the film 
as it enters and leaves the sprocket, and the 
shape of tooth required for sprockets of differ- 
ent diameters. These standards were 
Z22.6-1941 and Z22.18-1941, respectively. 
Sectional Committee Z22 of the American 
Standards Association has affirmed this action. 

There will be no American Standards to 
replace these two standards for some time. 
However, the Committee on Standards has 
concluded that the valuable information in 
the Chandler, Lyman, and Martin paper 1 
will lead to sprocket designs giving superior 
performance. This material provides, for any 
application, a flexible means of designing 
sprockets that will transfer the load of the 
film smoothly from one tooth to the next, 
thereby materially lengthening the life of the 
film and decreasing flutter. 

There are several reasons why this informa- 
tion is not ready for approval as an American 
Standard. These include: 

(1) The committee has emphasized that two 
of the dimensions, K and B, which determine 
the final shape of the tooth, depend on the 
outcome of extensive tests to show which 
shape results in the longest film life. 

(2) Some manufacturers are reluctant to 
abandon the shape given in previous standards 

1 J. S. Chandler, D. F. Lyman, and L. R. Martin, Propo- 
sals for 16-mm and 8-mm sprocket standards. Jour. 
SMPE, vol. 48, pp. 483-520; June, 1947. 

or by their own designers because they have 
found it to be successful for the particular 
sizes of sprocket they are using. 
(3) There has been disagreement among the 
members of the committee about whether 
standards should be written merely to insure 
interchangeability of parts or whether they 
should be based on affording optimum per- 
formance. These particular proposals are 
definitely in the latter class. 

Some of the preceding viewpoints were ex- 
pressed clearly by E. W. Kellogg. 2 

In view of the foregoing, the Committee 
has authorized the following abridgement of 
the Chandler, Lyman, and Martin paper, 
with additional data on 8-mm sprockets, 
printed in a form suitable for inclusion in the 
SMPTE Standards binder and entitled "Rec- 
ommendations for 1 6-mm and 8-mm Sprocket 
Design." The purpose is to give the informa- 
tion a more official status than that of a paper 
in the JOURNAL. 


In October, 1945, Committee Z22 of the 
American Standards Association referred back 
to the Committee on Standards of the Society 
of Motion Picture Engineers, Standards 
Z22.6-1941 and Z22. 18-1 941 covering 16-mm 
and 8-mm film sprockets, respectively. They 
were returned with the suggestion that the 
substitution of formulas for the specific dimen- 
sions given in the original standards would 
afford the designer a more flexible means of 
meeting the requirements of each particular 
application. The Chairman of the Commit- 
tee on Standards appointed a subcommittee to 

*E. W. Kellogg, Discussion by letter, 
vol. 51, pp. 437-440; October, 1948. 

Jour. SMPE, 

APPROVED: For a year's trial and criticism by the SMPTE Standards Committee, February 1, 1950. 





SSo.-o'i-S'r.ao2^2-s|-Sg^^ 88" 

ll ill ^ hill Jil" 8 ^ ill -o~ 
lllflitPihg 11*1 ifls-Hi 


SSSQ S S tS(5 


G ^ fi r^ 
u Q *ij Q -C J1 









oo ~ 

? ? 1 

00 Cfl 





11 o &! 

fi _T? 

/ - 




^- T-I a, 









+ 1 

s 2 
T "8 












o^ -5 







J x 


X C 



00 j 
00 C 

S + 

o m 

i : 


11 ^ 

00 * 

?? ? 

o m 

1 o 
^ d 



:- ?? 

f + ?? 

1 m 

? s s 

1 u 

v rt 




O^ v 


d ^ 



" in" 



1 ^: 



CvT > 



II i 



1 Vi 

1 + ^; 

^5 rt 

r <=>' 


o? c 


^ G- 

in c 

-* J3- 

1 c 

c + 

* + c 

r-l HH 















1 d 

O *i 



+1* . 


. ^^ 




^^ 1 


S -^ 




f*^ 0"^ 



O r-, 

o ^ 



( N (rt 


o u 

^ do 





II d 



o ^ 




S " 

bq &H c 

5 ^ 

0, C 

5 ^ 6 


ft, C 

1 1 II 

^ r*^ i^ 

i 1 









*O "*-* 

1 il 





00 TO 

00 -^ 








N **"* 

^ oo 



+ 1 

?? ' 


4,) W 

-g + 1 

+ 1 

+ 1 


^ *"* 



" o 


o ^ 

*+- p^ 

' s 


d .SP 







i JS 






^ . 




+ | 





*i* ^ 









o S 




o -g . 






NN "*~*^ ^G 


.2 8 




^ S"s 


c3 cd 


S "T3 *- 


I 2 




to ^1* 

> ^ Sfc 



1 -*5 

2 g* 

3 o 
S 2 





Drive Sprocket 
P = 0.300(1 - S min ) 

r, _ P X JV A nn , 

> ^ 
\ d 




5 ' 
II 1 

JS Qs C 

3 ^ 
3 1. 





} ^ R 

s c 
Co c 

^ . C 


^ cO A 

1 1 
CJ ft, C 

1 S' 
1 1 

* 1 s 

1 fe " 


, j *1 
+ 11 

in DO 

s |s 
s sfe 
n j! 

05 <j c 







prepare new standards. * It was for that sub- 
committee that these recommendations were 

Figures 1 and 2 are the recommendations 
for 16-mm and 8-mm sprockets, respectively. 
Provision has been made for camera, printer, 
and projector sprockets having any practi- 
cable number of teeth. Particular attention 
has been given to the shape of the film path 
and to the lateral profile of the sprocket itself 
and also to the lateral profiles of guides, rollers, 
and film gates. 

For simplicity of illustration, the film is often 
shown entering and leaving the sprocket in 
straight lines tangent to the root diameter, 
but by far the more usual path is a curve, 
either away from the sprocket or toward the 
sprocket. A minimum of y inch is 
proposed for R it the radius of the path that 
curves away from the sprocket. The pro- 
posed value of 0.7D for R 2 , the minimum 
radius when the film is curved toward the 
sprocket, is derived analytically. 

Accommodation for changes in film caused 
by shrinkage is the principal factor in the de- 
sign of sprockets. There are two reasons why 
this accommodation is necessary: (1) the 
film must riot be damaged prematurely by the 
sprocket (this is important for projection 
equipment in which the same film may be 
run many times), and (2) on sound and print- 
ing sprockets, the film must run at a relatively 
constant velocity in order to ensure freedom 
from flutter. Fortunately, it is possible to 
design for good results in both respects. 

Three aspects of sprocket design for which 
the potential shrinkage of the film must be 
taken into account are: (1) the circular pitch 
of the teeth, (2) the shape and thickness of the 
teeth, and (3) the lateral profile of the 

Circular Pitch of the Teeth 
' The circular pitch of the sprocket teeth can 
be made longer or shorter than the pitch of 
the film. Only in rare cases will the pitches 
be equal. The choice depends largely upon 

* O. Sandvik (Chairman), H. Barnett, J. A. Maurer, L. T. 
Sachtleben, and M. G. Townsley. 

the type of service expected from the sprocket. 
Types of Sprockets. In motion picture equip- 
ment there are two basic types of sprockets. 
With the first type, the film is urged forward 
against backward tension, and the sprocket 
is a drive sprocket. . The second type is the 
take-up or holdback sprocket. Here both the 
motion and the tension are forward; hence the 
film is held back by the sprocket. 

The important rules for drive and hold- 
back sprockets are: 

(1) A properly designed drive sprocket should 
have a circular pitch equal to or greater than the 
pitch of the film. 

(2) A properly designed holdback sprocket 
should have a circular pitch equal to or less than 
the pitch of the film. 

These rules, upon which the formulas of the 
recommendations are based, are so chosen 
that all slippage between the film and the 
sprocket is in the same direction as the tension 
on the film. Thus any friction between the 
film and the base of the sprocket serves to 
assist in the functioning of the sprocket rather 
than to increase the load between the film 
and the teeth. Also according to the rules, 
the leaving tooth is the one that does the work 
of driving the film or holding it back. There 
is, therefore, clearance between the entering 
tooth and its mating perforation, as long as 
the thickness of the teeth is such as to avoid 

In addition, there is the combination 
sprocket, which is often used in reversible 
apparatus where the function of the sprocket 
changes as the direction of motion changes. 
Also, in many cameras and in some projec- 
tors, one section of a single sprocket serves 
as a drive sprocket and another section as a 
holdback sprocket. Moreover, the function 
of a sprocket may change owing to the vary- 
ing tension exerted by the take-up, or for 
other reasons. Combination sprockets are 
not recommended for precision apparatus 
such as printers or other professional equip- 

The optimum pitch for a combination 
sprocket is a compromise, at best. If a prop- 




erly designed drive sprocket is used as a 
combination sprocket, all film operating under 
holdback conditions is forced against the 
direction of external tension by each tooth as 
it enters the perforation. The same is true of a 
holdback sprocket operating under drive 
conditions. Experience has shown that un- 
less special attention is given to guiding the 
film as it engages the sprocket, a drive sprocket 
makes a poorer combination sprocket than 
does a holdback sprocket. Therefore, the 
combination sprocket was designed to favor 
its action when it is driving film; its pitch 
matches that of film having a shrinkage equal 
to the minimum shrinkage plus one-third 
(instead of plus one-half) the shrinkage range 
for which accommodation is being made. 

Shape and Thickness of Sprocket Teeth 
Importance of Shape of Tooth. In all cases ex- 
cept that of perfect mesh there must be some 
sliding of the edge of the perforation up or 
down the face of the tooth. The shape of the 
tooth is important not only from the stand- 
point of wear of the film at the point of con- 
tact, but also as it relates to the sliding of the 
film along the root circle of the sprocket and 
to the manner of transfer of the load from one 
tooth to the next. In the worst case, all the 
shrinkage differential is absorbed by a sudden 
jump of the film as it leaves the tip of one tooth 
and comes into contact with the next tooth. 


These recommendations for film sprocket de- 
sign have been developed to give the design 
engineer an opportunity to specify sprocket di- 
mensions for particular applications and condi- 
tions. They consist of a number of simple 
formulas for the computation of tooth thickness, 
tooth shape, and circular pitch based upon the 
range of film shrinkage to be accommodated 
and the amount of contact between film and 
sprocket. The root diameter on which the 
film will run is computed from the circular 
pitch and the number of teeth on one end of the 

In cases where the film pitch does not match 
the sprocket pitch, the formulas for the sprocket 
pitch are such that the slippage of the film will 
be in the direction of external tension on the 

film, backward on feed or drive sprockets, and 
forward on holdback sprockets. 

The optimum pitch of the combination 
sprocket has been established to specify a 
sprocket that meshes perfectly with film having 
a shrinkage of the minimum shrinkage plus one 
third of the shrinkage range. Film of less 
shrinkage is forced against the direction of the 
external tension when this sprocket is operating 
as a drive sprocket. Film of greater shrinkage 
is forced against the direction of external tension 
when this sprocket is operating as a holdback 
sprocket. Combination sprockets should be 
avoided wherever possible. 

It will be noted that formulas for the maxi- 
mum permissible tooth thickness involve the use 
of fractions of a film pitch length. (See defini- 
tions for H and F.) This provides clearance for 
the teeth in partial engagement with the film. 

The shape determined by dimensions K and 
B provides clearance at the tip of the leaving 
tooth for paths between R\ and /? 2 . For sound 
and printing sprockets optimum conditions of 
flutter are obtained when the film path curves 
toward the sprocket and approaches the limit 
defined by /? 2 . A more precise formula for R* 
for sound and printing sprockets is l//? 2 = 
19/JV - 26 IN* - 0.060. The sprocket in this 
case should be designed and used as a drive 
sprocket. If it is necessary to use a sound 
sprocket as a holdback sprocket, it should be 
designed as a drive sprocket, and film guides 
must be provided to force the film onto the teeth. 

The dimensions shown in the lateral profile 
views provide clearance for film with lateral 
shrinkage from to 1.0 per cent for cameras 
and from to 1.8 per cent for projectors. Film 
of greater shrinkage can be used on these 
sprockets, but either the film will pull away 
from the guide or the fillet of the perforation 
will engage with the tooth. These lateral dimen- 
sions are applicable also to film gates, guides, 
and pull-down claws. 

The choice by the engineer of the range of 
film shrinkage to be accommodated must de- 
pend upon the experience of the manufacturer 
and the type of equipment being designed. 
Based on current film conditions, suggested 
ranges would be from to 1 per cent shrinkage 
for equipment using unexposed film and from 
to 1.5 per cent shrinkage for equipment using 
processed film. 

In most cases, particularly when tensions 
greater than 2 oz must be overcome, it is de- 
sirable to make H at least two film pitch lengths. 
An exception to this occurs when the film is 
positively guided through a path curving toward 
the sprocket. In this case, if the value of F is 
high at least 4 the value of H may be zero. 




&0.22 .J!< 
C G 


5 o 

u 5^.^ 

o G 


W 03 


G <N 2 





s-s-^s i-ai a 
s|sl 111I 

^-S.aS Ssl.sSS 


-M ^ 






I" 8 ! 3 



' JS s o 

^ <u o JJ 


be 6 "5 ^ 
G rt^ G 


-^^^ rt 
- ^ bo 


. 5OT3 


- ca 
"0 S 


CX s-i || G 

u .h 


S-5^ | 


03 u 



> *i S 

-M "C 

S-S J^ 
u o-Q 

H. oj 

>H ** M i ^^> 

(L) O V & O OJ 

e 2 -a G 8-s 

8 &.a .S 



























+ i 












o d 


OO l 


2 d 

OJ .. 

no o 

o ~r 

on o 

o -T ^ oo 4- 

U" SS ?? S SS ;? ^ 

iC oo ^ 

S fci ? ~-v 

-i ^"^ ^^ oo 

00 T 


i O 

a -a 

+ ' u 

? 2- 

^ s. 

*" ' L 

' S 

oj c 




^_ s ^ 


y ro 52 ""! 

t^- O in 


.B in 



, + 

<-; II o 
d o + 


1 CO' 
1 1 


11 ^ 2 " S ' ^ 

g 8 m 
8 i 


1 7 






d ^ 


o ^ 


r-i C^l 




as .5 



te "^ 





N ln 


k] o'o 

^j m m 


T-I ^ 





o ^ 

^ do 








II 1 



II 1 

a, c 


II 1 

a* c 


^ K 




.} S 


2 "^ 

<o no 



^> 00 




_C In 

C => 









lT 05 

^ 00 





+ 1 

+ 1 


'C c 

S + 




O *" 
















rs O 















^4 * 












o o 




'M fj 



2 s 











3 t~ 

S ^ 

o ^ 
o "o 



"3 "o 


GO C-> ^ 

^ -VJ ^ P 

g ^ 

-7 6 - 
! * i o 

^ T-^^. m 

Co C 

if c 

3 1 





g 4- x 

^ ST oa< 
^ .S u i 



a d s 
S II 1 




-S M | 
5 II 1 




o ~ , ^ 
k, Q a., q K 

&3 ft, Q K? 

Ii, = 

5 t- 1 


fc I cL |- 

^J^ ^ to u, ~ O 

I^!5 12 




It will be obvious from a study of the nature 
of the action of the tooth that the drive 
sprocket is the most critical with respect to 
the shape of the tooth. 

Epicycloid Curve. The most logical starting 
place for the analysis of the shape of the tooth 
is the curve generated by a point on the film 
relative to the sprocket when the film moves 
along its path without slipping on the root 
circle of the sprocket. Since we have as- 
sumed that the path is an arc of a circle, the 
curve so generated is an epicycloid. In the 
case of a straight path, the resulting curve is 
an involute and can be treated as a special 
case of the epicycloid in which the generating 
circle has an infinite radius. The generated 
curve is an epicycloid whether the generating 
circle curves away from the sprocket or curves 
toward the sprocket and actually encloses it. 
(Equations to aid in plotting the epicycloid 
are given in the complete paper.) 

Since the epicycloid is generated by a 
circle rolling on the root circle, it is the locus, 
relative to the sprocket, of a point on the film, 
such as the edge of a perforation, provided 
there is no slippage between the film and the 
sprocket. The epicycloid curve is a valuable 
reference from which the desired shape of the 
tooth can be deduced by proper allowance for 
the amount of film shrinkage to be accom- 

It is obvious that the reference epicycloid 
to consider is the one that corresponds to the 
most limiting condition, namely, the film 
path that curves away from the sprocket 
along the arc with the minimum radius. 

Allowance for Shrinkage at Tip of Tooth. Figure 
3 shows on a large scale the film and one tooth 
of a 1 2-tooth sprocket for 1 6-mm film. If the 
tooth is moving to the left and the film ten- 
sion is to the right, we have a drive sprocket. 
The circular pitch of the sprocket is greater 
than the pitch of the film. 

If the shape of the tooth is such as to guide 
the film along the epicycloid (Fig. 3, curve 1) 
the film will not slip on the sprocket until 

after it leaves the tip of the tooth, whereupon 
it will jump to the right and will stop suddenly 
when the next perforation engages the next 
tooth to the right. A tooth shape falling to 
the right of the epicycloid will let the film slip 
gradually and thus accommodate part of, or 
all, the shrinkage differential before the film 
reaches the tip of the tooth. The optimum 
condition is reached when the tooth just 
allows accommodation of the maximum 
shrinkage differential when the film is ready to 
leave the tip of the tooth. If the film is shrunk 
less, there will be full accommodation earlier 
and the film will leave the tooth before it 
reaches the top. 

D= 1.1400 INCHES 
K= 0.0804 INCH 

B= 0.0146 INCH 

FOR 1.5% 






0.004 0.008 0.012 0.016 0.020 0.024 0.028 

Fig. 3. Tooth Shape for 12-Tooth, 16- 
mm Sprocket. 

The shrinkage differential with which we 
are concerned is that for one pitch length of 
film. For example, if the shrinkage range is 
from to 1.5 per cent, the circular pitch of the 
sprocket is chosen to match unshrunk film 
(0.300 in.) and the maximum pitch differen- 
tial is 1.5 per cent of 0.300 in., or 0.0045 in. 
The proposed tolerances for the pitch of the 
drive sprocket are plus 0.0003 in., minus 
0.0000. Therefore, an additional allowance 
of approximately 0.0003 in. is made in estab- 
lishing the location of the tip of the tooth. 




Figure 3 shows two film positions, one in 
solid lines for a shrinkage of 1.5 per cent and 
one shown by broken lines for a shrinkage of 1 
per cent. In each case the lower edge of the 
perforation is 0.026 in. above the root circle, 
and the film is just ready to leave the tooth. 

Specification of Shape of Tooth. So far, only two 
points on the profile of the tooth have been 
located, one at the maximum working height 
of 0.026 in. and the other at the intersection 
with the root circle. Obviously, the manner 
in which the film is allowed to slip to take care 
of the shrinkage differential is controlled by 
the shape of the tooth between these two 
points. However, the relationship between 
the way the film slips and the running life of 
the film is not directly evident. The ultimate 
solution lies in exhaustive wear tests, with due 
consideration to all the other factors involved. 
A convenient method of specifying the shape 
of the tooth is to state the radius A* of a cir- 
cular arc and the distance B from the root 
circle to the center of the arc. (When B is 
positive, the center is inside the root circle.) 
This method is justified not only for its con- 
venience but for practical considerations of 
manufacture. That the circular arc is ade- 
quate can be seen in Fig. 3 from the close 
agreement of the circular arc with the epicy- 
cloid. One logical procedure for defining 
the shape of the tooth has been completely 
worked out and is described here, followed by 
a brief discussion of an alternative procedure. 

The first method is based on the circular 
arc that best approximates the epicycloid. 
As the shrinkage differential increases, the 
radius K of the tooth (Fig. 1) remains constant, 
but B, the distance from the root circle to the 
center of the radius, is increased. This brings 
the upper end of the tooth to the proper ter- 
minal point and provides a uniform shrinkage 
adjustment as the film moves up the face of 
the tooth. 

It may be argued that the above procedure 
will result in a tooth that slants too much at 
its base for good driving action, particularly 
when the tension on the film is high. The 

alternative procedure for determining the 
shape of the tooth overcomes this objection. 
By this method, the value of B is made inde- 
pendent of the range of shrinkage and ap- 
proximates the value given by the above 
equation for zero range of shrinkage. The 
value of K must then vary with the range of 
shrinkage, from about the value given above 
to lower values as more shrinkage is accom- 
modated. The resulting tooth is very nearly 
tangent to the epicycloid at its base but never 
crosses to the left of the epicycloid. This gives 
the steepest permissible tooth at the base. 
Equations for this procedure have not been 
completely derived, but the shape for 1.5 per 
cent shrinkage on a 1 2-tooth, 1 6-mm sprocket 
is shown by a light line on Fig. 3. The run- 
ning life of the film should be considered the 
most important criterion for the final choice. 
From the standpoint of flutter, there appears 
to be an advantage in the first procedure. 

It is also possible to specify an involute 
(with a given pressure angle) for the shape of 
the tooth. This is not recommended because 
it results in a tooth which is even more slanting 
at its base than the tooth obtained by the first 

Thickness of the Tooth at the Base. For the 
properly designed drive or holdback sprocket, 
the action of the tooth takes place at the end 
of the film path, where the tooth is leaving. 
For the cases of maximum differential be- 
tween the pitch of the sprocket and that of the 
film, there may be interference at the entering 
tooth (at the outside face for drive sprockets 
and at the inside face for holdback sprockets) 
if proper attention is not given to the selection 
of the thickness of the tooth. 

As given in the recommendations, the equa- 
tions apply to the first procedure for deter- 
mining the shape. The formula for the thick- 
ness of the tooth includes the following reduc- 
tions from the full longitudinal dimension of 
the perforation: one for the shrinkage of the 
film in the arc of contact; a second for the 
teeth in partial engagement; a third for the 
pitch tolerance allowed for the sprocket; and 




a fourth for positive clearance. If the tooth 
becomes too thin to provide the proper work- 
ing height or the desired total height, the arc 
of contact or the arc of partial engagement, or 
both, must be shortened. 

Other Considerations. One advantage of the 
equation form of the recommendations is its 
great flexibility. Suitable changes can readily 
be introduced to allow for the effects of stretch 
and distortion of the film. No attempt has 
been made in the present analysis to incor- 
porate these alterations. It is felt that more 
precise quantitative information is needed 
and that any such proposed changes should 
be backed by thorough testing before their 
presentation. On the basis of test observa- 
tions, for sprockets made according to the 
formulas of Fig. 1 , it is recommended that the 
tension not exceed 4 oz for drive sprockets nor 
8 oz for holdback sprockets. 

The number of film pitch lengths in the arc 
of contact and in the arc of engagement has a 
bearing upon the operation of the sprocket, 
especially if appreciable amounts of distor- 
tion are present. For practical reasons the 
minimum recommended value of H has, 
therefore, been set at 2, except for the special 
conditions noted in the Appendix of Fig. 1 . 

Lateral Profile 

Need of Specifying Lateral Profile. It is neces- 
sary to pay special attention to the lateral 
profile of sprockets in order to prevent the 
corners of the teeth from damaging the fillet 
of the perforation. Other zones to protect 
are the picture area and the sound-track area. 
The purpose, therefore, of specifying the lateral 
profile is to ensure that the tooth will be of the 
correct size and that it will be located properly 
in relation to the guide for the edge of the film. 
Moreover, it is necessary to locate the zones 
that are recessed or undercut below the root 
diameter so that the picture and sound-track 
areas will be protected. 

Edge to Be Guided. Several conditions must be 
established before the dimensions can be deter- 

mined. The first of these is the choice of the 
edge of the film to be guided. Figure 1 is 
based on a fixed guide at the sound-track 
edge of the film. This results in rails of ade- 
quate width at that edge, but it is necessary 
to restrict the lateral width of the tooth some- 
what because the tooth is so far from the fixed 
guide. On the other hand, if the fixed guide 
were placed at the perforated edge of the film, 
the lateral width of the tooth could be made 
greater, but the rails would be extremely 
narrow, if they could be provided at all. 
Channel guiding is subject to the disadvan- 
tages of both systems in that the tooth must be 
narrow and it is almost impossible to specify 
guiding rails that will not scratch the sound 
track. Detailed tables for the three methods 
of guiding, and for sprockets with two rows of 
teeth, are given in the complete paper. 


The principal advantages of these recom- 
mendations are their adaptability and their 
flexibility. They are adaptable to any 
application regardless of the size and the func- 
tion of the sprocket and also of the path and 
the shrinkage of the film. They are flexible 
because they are presented in such form that 
if changes are made in the physical properties 
of film or if research discovers new conditions 
of improved operation, the formulas can be 
adjusted to keep the recommendations up to 

NO TE: G. F. Vilbrandt supervised actual 
running tests with different types of sprockets 
and reported his preliminary findings. 3 His 
report is of interest for two reasons: (1) it 
describes the action of the various sprockets 
with films of different pitches; and (2) it 
indicates a method of testing that correlates 
the life of the film with the differential pitch, 
and with the tension to which the film is sub- 
jected. Further testing of this kind should 
indicate the tooth shape that will provide the 
longest film life. 

3 C. F. Vilbrandt, The projection life of 16-mm film. Jour. 
SMPE, vol. 48, pp. 521-542; June, 1947. 

Proposed American Standard 
16- Mm Projection Reels 

THE PROPOSED STANDARD which appears on the following pages was 
prepared by the 16- and 8-Mm Motion Pictures Committee, un- 
der the Chairmanship of Mr. Henry Hood. Completely new in 
appearance, it replaces the previous American Standard for 16-Mm 
Projections Reels Z22. 11-1941 and is being published here in the form 
of a proposal for ninety days trial and criticism. Engineers or equip- 
ment designers are invited to send their comments on the proposal to 
William H. Deacy, Jr., Staff Engineer at Society Headquarters, and 
are asked to do so before June 1, 1950. 

The original 16-Mm Reel Standard, Z22.ll, was found to be inade- 
quate during the war and at that time an American War Standard, 
Z52.33-1945, was developed to provide the armed services with a 
more detailed set of specifications. At the end of the war, a subcom- 
mittee, under the chairmanship of Mr. D. F. Lyman, examined both 
the prewar and the wartime standards, with a view toward combining 
the best parts of each in a form that would be most useful to equip- 
ment designers and manufacturers. The project was subsequently 
transferred to the 16- and 8-Mm Motion Pictures Committee and this 
proposal is the formal recommendation of that Committee. 

In developing this proposed standard, the Committee found that 
some of the dimensions and tolerances of the wartime standard were 
so rigid as to be commercially impractical and that it was necessary 
to take a more realistic point of view, since reels had to be manufac- 
tured in quantity and sold at a reasonable price. These considera- 
tions account in particular for the wide tolerances shown on the out- 
side and core diameters. As indicated in the note following Table 2, 
the Committee has expressed a hope that reel manufacturers will 
adopt the dimensions recommended whenever they find it necessary 
to manufacture new production tools. 

The only question not completely resolved in the many committee 
meetings required to develop this new proposal had to do with the 
shape of the spindle hole. One group favored the use of square holes 
in both flanges while another group recommended one square and one 
round hole. Consideration was also given to the possible alternative 
of using a keyway as defined by Dimensions U and V, that could be 
added to the round hole. The Committee feels that comments on 
this particular point would be very helpful. 





Proposed American Standard 

16-Millimeter Motion Picture Z22U 

Projection Reels 


.^* ^s. 

' /^^ 

x^ "x 


/ ^_ \ 

.-W-* -AT CORE 

/ /*' "*^ \ 


/ ^ VI 


1/1 n i 1 c i 



i 1 ^^> /' i 


\ X s \ } 

( r 8 ^ 

\ / 


t / u 

-T i /' / 

^v^ / 

** *T ' > r 

\^ ^/ 

\ '' * 






Dimension Inches 


A 0.319 

+0.000 fi . n +0.00 
-0.003 u -0.08 

B 0.319 

+0.000 oin +0.00 
-0.003 U -0.08 

W, at periphery ' 0.660 

+0.045 1A7A+ 1 - 14 

-0.025 1676 -0.64 

at core * 0.660 

0.010 16.76 0.25 

at spindle holes 0.660 0.015 16.76 0.38 

T (exclusive of 0.027 

minimum 0.69 minimum 

embossing) 0.066 

maximum 1.68 maximum 

S at periphery 3 

(including flared, QA9 
rolled, or beveled 

24.43 maximum 


U 0.312 

0.016 7.92 0.41 

V 0.125 

+0.005 , lft +0.13 
-0.000 -0.00 

Flange and core Q Q31 
concentricity 4 


See Notes on p. 3. 





Proposed American Standard 

16-Millimeter Motion Picture 
Projection Reels 



P. 2 of 3 pp 

Capacity Dimension Inches Millimeters 

Capacity Dimension 

Inches Millimeters 

200 Feet 5 D, nominal 5.000 127.00 

1200 Feet D, nominal 

12.250 311.15 

(61 Meters) maximum 5.031 127.79 

(366 Meters) maximum 

12.250 311.15 

minimum 5.000 127.00 


12.125* 307.98* 

C, nominal 1.750 44.45 

C, nominal 

4.875 123.83 

maximum 2.000* 50.80* 


4.875 123.83 

minimum 1 .750 44.45 


4.625* 117.48* 

Lateral run- - . .. 
out, 6 maximum - 570 ] ' 45 

Lateral run- 
out/' maximum 

0.140 3.56 

400 Feet 5 D, nominal 7.000 177.80 

1600 Feet D, nominal 

13.750 349.25 

(122 Meters) maximum 7.031 178.59 

(488 Meters) maximum 

14.000* 355.60* 

minimum 7.000 177.80 


13.750 349.25 

C, nominal 2.500 63.50 

C, nominal 

4.875 123.83 

maximum 2.500 63.50 


4.875 123.83 

minimum 1.750* 44.45* 


4.625* 117.48* 

/L m - 8 2.03 

Lateral run- 
out, 6 maximum 

0.160 4.06 

800 Feet D, nominal 10.500 266.70 

2000 Feet D, nominal 

15.000 381.00 

(244 Meters) maximum 10.531 267.49 

(610 Meters) maximum 

15.031 381.79 

minimum 10.500 266.70 


15.000 381.00 

C, nominal 4.875 123.83 

C, nominal 

4.625 117.48 

maximum 4.875 123.83 


4.875 123.83 

minimum 4.500* 114.30* 


4.625 117.48 

Lateral run- mon i n 
out, 6 maximum ai20 3 ' 05 

Lateral run- 
out, 6 maximum 

0.171 4.34 




Proposed American Standard 

16-Millimeter Motion Picture 
Projection Reels 



* When new reels are designed, or when new tools 
are made for present reels, the cores and flanges 
should be made to conform, as closely as practicable, 
to the nominal values in the above table. It is hoped 
that in some future revision of this standard the as- 
terisked values may be omitted. 

1 Except at embossings, rolled edges, and rounded 
corners, the limits shown here shall not be exceeded 
at the periphery of the flanges, nor at any other dis- 
tance from the center of the reel. 

- If spring fingers are used to engage the edges 
of the film, dimension W shall be measured between 
the fingers when they are pressed outward to the 
limit of their operating range. 

{ Rivets or other fastening members shall not ex- 
tend beyond the surfaces of the flanges more than 
1/32 inch (0.79 millimeters). 

4 This concentricity is with respect to the center line 
of the hole for the spindles. 

5 This reel should not be used as a take-up reel on 
a sound projector unless there is special provision to 
keep the take-up tension within the desirable range 
of l'/2 to 5 ounces. 

6 Lateral runout is the maximum excursion of any 
point on the flange from the intended plane of rota- 
tion of that point when the reel is rotated on an accu- 
rate, tightly fitted shaft. 


Dimensions A and B were chosen to give 
sufficient clearance between the reels and the 
largest spindles normally used on 16-milli- 
meter projectors. While some users prefer a 
square hole in both flanges for laboratory 
work, it is recommended that such reels be ob- 
tained on special order. If both flanges have 
square holes, and if the respective sides of the 
squares are parallel, the reel will not be suit- 
able for use on some spindles. This is true if 
the spindle has a shoulder against which the 
outer flange is stopped for lateral positioning 
of the reel. But the objection does not apply if 
the two squares are oriented so that their re- 
spective sides are at an angle. 

For regular projection, however, a reel with 
a round hole in one flange is generally pre- 
ferred. With it the projectionist can tell at a 
glance whether or not the film needs rewind- 

P. 3 of 3 pp. 

ing. Furthermore, this type of reel helps the 
projectionist place the film correctly on the 
projector and thread it so that the picture is 
properly oriented with respect to rights and 

The nominal value for W was chosen to 
provide proper lateral clearance for the film, 
which has a maximum width of 0.630 inch. Yet 
the channel is narrow enough so that the film 
cannot wander laterally too much as it is 
coiled; if the channel is too wide, it is likely to 
cause loose winding and excessively large 
rolls. The tolerances for W vary. At the core 
they are least because it is possible to control 
the distance fairly easily in that zone. At the 
holes for the spindles they are somewhat 
larger to allow for slight buckling of the 
flanges between the core and the holes. At 
the periphery the tolerances are still greater 
because it is difficult to maintain the distance 
with such accuracy. 

Minimum and maximum values for T, the 
thickness of the flanges, were chosen to per- 
mit the use of various materials. 

The opening in the corner of the square 
hole, to which dimensions U and V apply, is 
provided for the spindles of 35-millimeter re- 
winds, which are used in some laboratories. 

D, the outside diameter of the flanges, was 
made as large as permitted by past practice 
in the design of projectors, containers for the 
reels, rewinds, and similar equipment. This 
was done so that the values of C could be 
made as great as possible. Then there is less 
variation, throughout the projection of a roll, 
in the tension to which the film is subjected by 
the take-up mechanism, especially if a con- 
stant-torque device is used. Thus it is necessary 
to keep the ratio of flange diameter to core 
diameter as small as possible, and also to 
eliminate as many small cores as possible. For 
the cores, rather widely separated limits (not 
intended to be manufacturing tolerances) are 
given in order to permit the use of current reels 
that are known to give satisfactory results. 


A Restatement of Policy 

The Society of Motion Picture and Television Engineers is com- 
posed of technical representatives from all phases of the industries 
involved in the reproduction of scenes with motion, many of whom 
compete most vigorously with one another in the open markets. It is a 
matter of considerable pride to the Society that such vigorous and 
sometimes bitter competitors in commercial life find it profitable to 
meet together under the impartial auspices provided by the technical 
sessions and the committee activities of the Society. The impartial- 
ity of these meeting grounds must never be questioned if the Society 
is to enjoy the support of all its members, and if it is to grow in ac- 
cordance with its opportunities. 

It is the purpose of this restatement to emphasize a fundamental 
policy of Society operation, which is directed toward the maintenance 
of such an unquestioned impartiality. Briefly, this policy is simply 
that the Society shall refrain altogether from participation in the 
comparative testing of competitive goods. 

The obvious wisdom of such a procedure would scarcely seem to re- 
quire a formal statement such as this. However, opportunities for 
transgression continually arise, particularly in the conduct of the 
normal affairs of the Engineering Committees of the Society. The 
development of suitable test methods, and, in many cases, the estab- 
lishment of the manufacturing tolerances and performance levels 
which can be expected of good equipment are among the most im- 
portant duties of these Committees. To develop new methods and 
limits, the full co-operation of industry is relied upon, including the 
opportunity to make extended tests on commercial equipments, with 
mutual confidence that the results of these tests will not be made the 
basis for competitive publicity involving the Society. 

The Society, of course, encourages the use by industry of approved 
Society methods and tolerances in competitive commercial testing. 
Moreover, if a commercial interest wishes to publicize the fact that 
Society-approved methods were used to authenticate its claims, that, 
too, is encouraged. Authorization is never granted, however, for any 
implication of Society participation in, or validation of, the test re- 
sults themselves. 


Board of Governors 

The first meeting of the Board of Governors in 1950 was held in 
New York, January 31, with Earl I. Sponable presiding. As is cus- 
tomary, the agenda was a heavy one and the meeting ran well into 
the late afternoon. The Board reviewed the Society's accomplish- 
ments for the previous twelve months and studied the details of all 
business operations involved in turning out the JOURNAL and test 
films, of engineering and non-technical committee work, as well as 
administration of Society membership matters. The resulting 
year-end fiscal picture was analyzed in detail and together with 
recommendations of the Officers, Headquarters Staff, and many 
committees, was used as a basis for planning work to be undertaken 
in the coming year. 

William B. Lodge, whose term as a Governor expired in Decem- 
ber, was reappointed for an additional year to fill the second half 
of a two-year term vacated when Fred T. Bowditch was elected 
to the office of Engineering Vice-President. As Vice-President in 
Charge of Engineering for Columbia Broadcasting System, Mr. 
Lodge brings the views and opinions of television engineers gener- 
ally to deliberations of the Board. 

Among the changes approved in the recent amendment to the 
Society's Constitution was the creation of two additional Gover- 
norships, increasing the number of Board members to twenty-four. 
There will now be twelve elected Governors, four from the Eastern 
time zone, four from the Central Zone and four from the Mountain 
and Pacific zones. After some consideration the Board ruled that 
it would decide at the April meeting whether the two new offices 
should be filled temporarily by appointment, or by the 1950 election. 
Since the recently approved Constitutional Amendment is now 
in effect, the President's Committee on Revision of the Constitution 
and Bylaws has turned its attention toward drafting a proposed 
Amendment to the Bylaws. Changes which the Committee was 
asked to consider were intended to bring both documents more 
nearly into agreement and it was hoped that certain complex word- 
ing, redundant phraseology, and some outright conflicts would be 
resolved. Prior to the meeting, the first formal draft of the Com- 
mittee's proposal was reviewed by the Board members who have now 
approved referring it to the voting membership for their considera- 
tion. This will involve publication in the March JOURNAL and dis- 
cussion followed by a formal vote during the 67th Convention. 

67th Semiannual]Convention 

On April 24, the 67th consecutive semiannual convention of the 
Society will get underway at the Drake Hotel in Chicago. Bill 
Kunzmann, Convention Vice-President, George W. Colburn, Cen- 
tral Section Chairman, and R. T. Van Niman, Chicago Vice-Chair- 
man of the Papers Committee, report that advance arrangements 
are shaping up nicely and they plan to provide for both a full pro- 
gram and large attendance. 

The five-day convention is now scheduled to include ten techni- 
cal sessions, with papers on a variety of subjects ranging from pro- 
jection arc lamps to production techniques for television studios. 
Members and their guests will start the ball rolling at 12:30, 
Monday, April 24, with a 'Get-Together' luncheon in the Gold 
Coast Room of the Drake Hotel. On Wednesday, Bill Kunzmann 
will, as usual, call time out from the official business of this, his 67th 
convention, for a midweek evening of fun and frolic. The 67th 
Banquet and Dance will follow a one-hour Cocktail Party, sched- 
uled for 6:45 P.M. in the hotel's French Room. 

Mrs. George W. Colburn, Convention Hostess, invites members 
to take their wives to Chicago and promises them a worth-while 
week because her committee is arranging a Ladies' Program that 
will be both interesting and entertaining. 

Bill Kunzmann announces appointment of the following Con- 
vention Committee Chairmen : 
Central Section and Local Arrangements, G. W. Colburn 
Papers Committee 

Chairman, N. L. Simmons, Jr. Vice-Chmn., Montreal, H. S. Walker 

Vice-Chmn. , Chicago, R. T. Van Niman Vice-Chmn. , New York, E. S. Seeley 

Vice-Chmn., Hollywood, L. D. Grignon Vice-Chmn., Washington, J. E. Aiken 
Publicity, Chairman, Harold Desfor 

Assisted by Leonard Bidwell and R. T. Van Niman 
Registration and Information, E. R. Geib 

Assisted by C. E. Heppberger, J. L. Wassell, and C. L. Lootens 
Luncheon and Banquet, Carrington H. Stone 
Hotel Reservations and Transportation, Harold A. Witt 
Membership and Subscription, Lee Jones 

Vice-Chairman, Central Section, A. H. Bolt 
Ladies' Reception Committee Hostess, Mrs. G. W. Colburn 
Public Address Equipment, Robert P. Burns 

Assisted by R. Hilton and R. Gray 
Projection Program, 35-mm, I. F. Jacobsen 

Assisted by members Chicago Projectionists Local 110, 1.A.T.S.E. 
Projection Program, 16-mm, H. H. Wilson 


The Papers Committee has sent Author's Forms to many mem- 
bers who expect to present technical papers. If you would also 
like to be on the program, don't fail to get your Author's Form from 
the member of the Papers Committee nearest you. His name and 
address are listed on p. 116 of the January JOURNAL. 

Society Announcements 

Membership Directory 1950 

To be certain that your name is listed correctly in the Society's Membership 
Directory for 1950, return your envelope-questionnaire promptly. 

Nominations 1950 

All voting members are invited to recommend candidates for the ten vacancies 
on the Board of Governors which will occur on December 31, 1950. Refer to p. 
113 of the January JOURNAL for the specific offices and governorships to be va- 
cated, as well as for the names and addresses of Nominating Committee members. 

Society Awards 1950 

Candidates are considered each year for five formal Society awards. A concise 
listing of the awards and the qualifications of recipients appeared on p. 113 of the 
JOURNAL for January, while a detailed reference with the names of previous 
recipients appeared in the April, 1949, JOURNAL. 

Test Film Catalog 

A new catalog, listing all test films for both motion picture and television use 
produced by the Society and the Motion Picture Research Council, has just come 
off the press. Copies are available to all who use 16- or 35-mm projectors or 
sound equipment. Members are urged to make copies available to their own 
business organizations and also to friends and acquaintances who are concerned 
with the design and manufacture of film handling equipment or with the quality 
reproduction of picture and sound. 

High-Speed Photography 

The Society is pleased to announce that High-Speed Photography, Volume 2, a 
complete reprint of the articles on high-speed photography from the November, 
1949, JOURNAL, is now available. This follows Volume 1, published as a supple- 
ment to the JOURNAL in March, 1949, and is of great value to all who use high- 
speed photographic methods of analysis. 

Seventeen articles by noted authorities cover the field of equipment now 
available commercially and in addition describe a number of special cameras and 
associated control equipment developed for unusual needs. Of these seventeen, 
five are primarily "technique" papers that give new users of photo-analysis 
methods a good grounding in various specific applications with an idea of the broad 
capabilities of scientific photography. 


Volume 1 (129 pp., $1.50) and Volume 2 (177 pp., $2.00) are companion refer- 
ence works developed by the High-Speed Photography Committee under the 
chairmanship of John H. Waddell. They are the only publication^ <f their kind 
and are therefore essential to the serious use of photography in government und 
industry research laboratories. 

Journals Out of Stock 

The Society's stock of JOURNAL issues for January, March, and July, 1949, has 
been exhausted as a result of an unexpected increase in demand and the Society's 
Headquarters is anxious to purchase a stock of each. Members or libraries hav- 
ing extra copies available are invited to send them in. The going price is 75c. 

Engineering Committees 

Theater Television and the FCC 

Frequency allocation continues as a topic of timely interest to 
members who have been following the work of our Theater Tele- 
vision Committee. Most recent sign of activity in this direction is 
a Public Notice from the Federal Communications Commission, 
released on January 11, and announcing plans for a Hearing on 
Allocation and Rule Making in the near future. This is by way of 
reply to statements by the Society and four industry groups favor- 
ing allocation of frequencies for theater television filed with the 
Commission in August, 1949, and to 26 individual petitions for a 
Public Hearing filed subsequently. 

The Commission in its January notice indicated that it desires 
"to obtain full information concerning all aspects of theater tele- 
vision; and to afford all interested persons an opportunity to par- 
ticipate in furnishing related information." 

The Commission reports that the Hearing would be held upon 
the following ten issues : 

(a) To determine whether the existing (c) To obtain full information con- 
and proposed transmission requirements cerning existing or proposed methods or 
for theater television can be satisfied by systems for exhibiting television pro- 
existing and proposed common carrier grams on large screens in motion pic- 
wire facilities or by existing and pro- ture theaters or elsewhere. 

posed common carrier fixed station fa- (d) To obtain full information con- 

cilities operated in bands of frequencies cerning existing or proposed methods 

now allocated to such stations. or systems for transmitting or relaying 

(b) To determine the order of fre- television programs from the point of 
quencies and the spectrum space re- pickup to the exhibiting theater, by 
quired, if any, at each order of fre- use of radio frequencies, coaxial cable, 
quency which would be necessary to wire, or other means, including intra- 
establish a theater television service. city and inter-city transmission. 


(e) To obtain full information concern- carriers for hire in interstate communi- 

ing any technical data obtained in cations by wire or radio, within the 

experimental operations conducted in meaning of Section 3(h) of the Com- 

the theater television field, or otherwise munications Act of 1934, as amended. 

(i) To determine whether, if fre- 


service, and commercial feasibility of 

the service. ( j ) In the light of the evidence adduced 

(g) To obtain full information con- under the foregoing issues, to deter- 

cerning plans or proposals looking mine whether or not the public interest 

toward the establishment of theater would be served by the issuance of a 

television on a commercial or non- proposal for allocation of frequencies to 

commercial basis. a theater television service and by the 

(h) To determine whether persons promulgation of proposed rules and 

engaged in furnishing theater television engineering standards governing such a 

services would be engaged as common service. 

All who desire to appear before the Commission on this question 
of the allocation of frequencies for a nation-wide theater television 
service have been invited to do so and are asked by the Commission 
to indicate their intentions by February 27. 

Society Recommendations 

In recent years many members who served on engineering committees of the 
Society have felt the need of some new form of official Society endorsement, short 
of formal standardization, that could be given to those reports or other committee 
conclusions that appear inappropriate for standardization under normal American 
Standards Association procedure. The Board of Governors recognized the need 
for validating these reports and in 1949 approved the publication of Society 
Recommendations. The Board reasoned that since study of technical problems by 
a Society engineering committee represents a great many man-hours and usually 
develops information of real technical value, details of certain projects that did 
not quite qualify for standardization should not be put away to collect dust or be 
submerged in annual progress reports but be available as published documents. 
It is the purpose of these Recommendations to formalize tentative conclusions so 
that they will be available for later engineering or research work. 

The Recommendations for 16- and 8-Mm Sprockets in this issue of the JOURNAL 
mark the first appearance of this new type of Society technical document. Two- 
column format will, for the present at least, distinguish this and forthcoming 
Recommendations from the rest of the JOURNAL. Following publication, they 
will be reprinted to fit the standards binder but will be on colored paper to set 
them apart from the familiar American Standards on Motion Pictures. 

Reprints of the Sprocket Recommendations will be available sometime after 
March 15th and all whose names appear on the Standards Mailing List will re- 
ceive order forms. Distribution is by no means restricted and all engineers who 
wish to have this document at hand for reference are urged to order promptly. 


35-Mm Sound Heads 

The theater equipment industry has long been plagued with 
serious projector and sound head interchangeability problems. The 
lack of formal standards for such important details as size and loca- 
tion of mounting holes or dimensions and speeds of projector drive 
gears forces each manufacturer to provide a complete series of 
adapter kits to permit matching his equipment to all other combina- 
tions of projectors, sound heads, bases, magazines, preview attach- 
ments, etc. Because the problem is such a complex one, any real 
standardization is many years away; but for the time being, the 
Film Projection Practice Committee, under the chairmanship of 
L. W. Davee has provided a measure of relief. They have assem- 
bled a combined reference file of the basic dimensions that affect 
interchangeability of 25 different types of 35-mm theater sound re- 
producers. Included in this "1949 Sound Head Survey" are: 

Ballantyne RSM-6 & RSM-8 RCA MI-9001 

Century R2 & R6 RCA MI-9030 Series, MI-9050 

Century RC & R5 Series, MI-9060 Series, 

International SH-1000 MI-9070 Series 

Projector through SH-1006 Wenzel WSH-3 

Motiograph SH-7500 Westrex Master R-2 

RCA MI-1040 Series, Westrex Standard R-3, 

MI-1050 Series Advanced R-4 

Copies have been sent to the manufacturers who participated in 
the survey and they are now made available for purchase at $10 for 
each set of eleven, 24 X 36 in. blueprints. 

New ASA Correlating Committee Formed 

A new Correlating Committee on Photography and Cinematography is 
being formed to supervise the work of Sectional Committees Z22 on Motion 
Pictures and Z38 on Photography of the American Standards Association. 
Believing this change would reduce the work load of Z38 projects as well as 
allow projects of both groups to flow more efficiently, the ASA suggested such 
a move several years ago. 

Approval was delayed by concern of several Z22 members who felt that 
their own efficiency would be reduced through additional approval steps re- 
quired. They were also concerned lest the word "cinematography" replac- 
ing "motion pictures" in the committee's title imply an undesirable limiting 
of the scope of projects they could undertake. 

Agreement was reached, however, when every assurance was given that, 
basically, proposed standards on motion pictures would be handled as before. 

The new Committee will have fifteen members. Two Society representa- 
tives and two from the Motion Picture Research Council will be appointed in 
February and will attend the first meeting, scheduled for late in March. De- 
tails of operation will be developed at that time and a report on the new 
organization will probably be presented at the Society's 67th Convention in 
Chicago during the last week in April. 


Section Meetings 

Atlantic Coast 

The Atlantic Coast Section meets on February 15, 7:30 P.M. at the Reeves 
Sound Studios, 304 East 44th St., New York City. Sound recording and tele- 
vision engineers are certain to be interested in the synchronous one-quarter inch 
magnetic tape recorder that will be described and demonstrated by Drs. D. G. C. 
Hare and W. D. Fling of Fairchild Recording Equipment Corp. This equipment 
is being used regularly by at least one New York television station for double- 
system television film recording. 


The February meeting of the Central Section is scheduled for 7:30 P.M., Febru- 
ary 17, in the 7th Floor Auditorium of the Western Society of Engineers, 84 E. 
Randolph St., Chicago. George W. Colburn, Section Chairman, reports that this 
will be a joint meeting with the Institute of Radio Engineers and the Chicago 
Audio and Acoustical group. There will be four technical papers, two on sound, 
one on television, and one on color. The first, "Some New Developments in the 
Field of Sound Reproduction," will be presented by Dr. Harry F. Olson, RCA 
Laboratories, Princeton, N.J., who will describe RCA's new highly directional 
microphone and its use in television and motion picture studios. The duo cone 
loudspeaker will also be described and a low cost, wide range recording and repro- 
ducing system will be demonstrated. 

The second paper, "Development and Application of the Short 16-In. Metal 
Kinescope," will be presented by Mr. Lloyd E. Swedlund of RCA, Lancaster, Pa. 
In addition to reviewing the history behind the development and design of the 
16GP4 picture tube, Mr. Swedlund will present practical information on its use 
and the use of several other new tubes. 

Mr. Frank Mclntosh of Mclntosh Engineering Laboratory, Silver Springs, 
Md., will present a paper, "New Developments in Audio Amplifiers." This 
amplifier has been much discussed in audio engineering circles and is certain to 
capture the interest of sound engineers. 

A representative of the duPont Research Laboratory at Parlin, N.J., will de- 
scribe briefly the new Palymar Color Process. 

Preceding the meeting, there will be a Speakers Dinner in the 5th Floor Dining 
Room of the Western Society of Engineers Building. 

Pacific Coast 

The Pacific Coast Section meeting for February is a double-header scheduled 
for the evening of February 14. One part is "Recent Problems and Develop- 
ments in Magnetic Film Recording," to be presented by Robert Herr, Research 
Physicist of Minnesota Mining and Manufacturing Co. All motion picture and 
television engineers who have been concerned with magnetic sound recording are 
acquainted with Mr. Herr and are certain to be interested in his discussion of some 
of the problems and techniques in the use of 35-mm magnetic films for motion 
picture sound work. This field of recording is distinct from applications which 
use non-synchronous tape. Aging characteristics of the magnetic medium are 
particularly important where sound records are to be used after a period of storage 


or where there is delay in shipment. The data which will be presented on the 
various effects will be of considerable benefit to studio sound engineers who are 
now converting from photographic recording. 

The other part has: first, "A New Technique for Synchronizing Multiple 
Television Originations" by Harold Jury, Chief Television Engineer, Don Lee 
Broadcasting System; and second, "Progress Report on an Electronic Background 
Projection System," by Wayne Johnson, KFI-TV, Los Angeles. Use of the pro- 
jected background captured the interest of television engineers many years ago as a 
means of saving production time and space in television studios. Mr. Johnson's 
talk will include descriptions of an electronic method of applying'a moving or still 
photographic background to a television picture. 

Book Reviews 

16-Mm Sound Motion Pictures, by William H. Offenhauser, Jr. 

Published (1949) by Interscience Publishers, Inc., 215 Fourth Ave., New York 3. 
546 pp. + 15 pp. index + 19 pp. appendix + xii pp. 123 illus. 6 X 9 in. Price, 

16-Mm Sound Motion Pictures- is described by the author as "a manual for the 
professional and the amateur." What makes it more than a manual, however, 
is the inclusion of a good deal of information derived from the practical experience 
of the author. 

Most of the chapters are devoted to the technical aspects of 16-mm photog- 
raphy, sound recording, editing, processing, and projection. Not only are repre- 
sentative equipments described with the aid of many illustrations, but also the 
techniques and methods are discussed at great length. Methods of quality con- 
trol of picture and sound, from the planning of a picture to the projection of its 
prints, are outlined. The problem of emulsion position and precautions to be 
taken in this regard throughout the steps of picture making are clearly described. 
There is a chapter on the history and growth of 16-mm film and its relation to 
other sizes. 

The chapter on film processing and printing is especially good. There is a 
chapter on color film and the duplicating process. Film in television is treated 
from the technical standpoint. Throughout the book, how-to-do-it informa- 
tion is profuse, in some cases, down to the last detail. Measurements in the proc- 
esses are carried through step by step, with the help of illustrative examples. 
This is very helpful and increases the value of the book. 

Liberal reference is made throughout the text to motion picture dimensional 
standards and practices of the American Standards Association, and their im- 
portance in attaining professional results is stressed. An appendix contains 
standards of nomenclature and various useful charts and symbols. 

The book is recommended to the many technicians working with 16-mm film. 
Each specialist not only will find it helpful in his particular area but also he will 
find that it enables him to broaden his technical knowledge in this rapidly ex- 
panding field. 


Warner Bros. Pictures 

Burbank, Calif. 


The Recording and Reproduction of Sound, by Oliver Read 

Published (1949) by Howard W. Sams & Co., Inc., Indianapolis 1. 358 pp. -f- 
6 pp. index + x pp. 256 illus. Price, $5.00. 

All those interested in the recording and reproduction of sound will discover 
in this book a wealth of information that will prove very useful, whether to them 
as hobbyists, operators of commercial equipment, or even as engineers. It de- 
scribes the subject in clearly understood language, yet uses less mathematics in 
doing so than has been observed by this reviewer in any other similar book. The 
engineer, who is frequently a specialist in but one or two methods of sound re- 
cording or reproduction, should also welcome in this book an opportunity to re- 
view in considerable detail the other sound recording and reproducing tech- 
niques; and, as a result, his broadened knowledge of the subject should prove 
helpful in his own more limited field. 

The chapter dealing with Columbia L.P. and RCA 45-rpm records is particu- 
larly timely and it points out the specifications and features of both systems as 
compared to the older 78-rpm records. 

Of particular interest to all is the last chapter which covers proposed NAB 
Recording and Reproducing Standards and proposed American Standard Acous- 
tical Terminology. 

The appendix contains numerous useful tables, charts, and formulas for elec- 
tronic service engineers as well as a tabulation of numerous disk recording troubles 
with their causes and cures. It also includes a glossary of many of the terms used 
primarily in the disk recording field. 


Radio Corporation of America 
. Hollywood 28, Calif. 

Meetings of Other Societies 

Institute of Radio Engineers, National Convention, March 6-9, New York, N. Y. 

Inter-Society Color Council, Annual Meeting, March 8, New York, N.Y. 

Optical Society of America, Winter Meeting, March 9-11, New York, N.Y. 

Armed Forces Communications Assn., Annual Meeting, 

May 12, New York and Long Island City 
May 13, Fort Monmouth, N. J. 

Institute of Radio Engineers, Technical Conference, May 3-5, Dayton, Ohio 

Acoustical Society of America, Spring Meeting, June 22-24, State College, Pa. 

SMPTE Officers and Committees: The names of Society Officers 
and of Committee Chairmen and Members are published annually in 
the April issue of the JOURNAL. Changes and corrections to these 
listings are published in the September JOURNAL. 


New Products 

Further information concerning the material described below can 
be obtained by writing direct to the manufacturers. As in the case 
of technical papers, publication of these news items does not consti- 
tute endorsement of the manufacturer's statements nor of his products. 

The new Huggins Ames Type A Mercury Arc Lamp has been designed to produce 
light intensities of 90,000 candles/sq cm. It is made by Huggins Laboratories, 
778 Hamilton Ave., Menlo Park, Calif. Shown in the illustration are the lamp- 
holder with lamp extracted and, in the background, the a-c power supply. Light 
output is 65 lumens/ watt; power consumption of the lamp is 2 kw (1.2 amp at 
1750 v). 

Arc dimensions are 2.85 cm (1.125 in.) by 1 mm (0.039 in.). Cooling is ac- 
complished with ordinary tap water, 2*/2 gpm being required. Alternatively, 
a closed-circuit distilled-water system can be used. Average life at rated maxi- 
mum brilliance is 5 hr, and appreciably more at reduced voltages. In the stand- 
ard model, 100 per cent intensity is reached at 4358 A, with an 80% peak at 5461 
A, and a 73% peak at 4047 A, with a maximum radiation of 0.08 w per steradian 
per Angstrom. Intermediate areas average approximately 35%. Quartz ac- 
cessories can be supplied for operation in the ultraviolet region. Direct-current, 
flash, and stroboscopic power supplies are reported as under development for 
special applications. 

Uses reported are in interferometers, Schlieren optical systems, shadowgraphs, 
monochromators, in high-speed photography, and high-powered ultraviolet 
sources in the production of chemical, biochemical, and ionization changes in 
substances under study or processing. 


Employment Service 


Project Engineer: Mechanical engi- 
neering graduate experienced in de- 
signing from specifications; optical 
instruments, precision cameras, me- 
chanical servo, and gear or 3-bar com- 
puters, analytical work in stress and 
vibration. R. A. Barbera, 663 Oving- 
ton Ave., Brooklyn 9, N.Y. 

TV and Motion Picture Engineer: 3 
yr experience in motion picture engi- 
neering and research at Philips Physi- 
cal Laboratories, Eindhoven; 6 yr as 
TV-Director, same firm; 3 yr as Di- 
rector of Decca plant in Belgium. De- 
sires asignment in any part U.S.A. 
Highest qualifications and references 
U.S. firms. Will visit New York and 
Chicago at the end of February. 
Write Fernand Beguin, c/o Mr. Marc 
Albanese, 416 Madison Ave., New York 

In technical phase: Motion picture or 
still photography. 4 yr experience in 
research, development, and testing, 
both color and b & w films. Gradu- 
ating from M.I.T. June, 1950. Mem- 
ber, SMPTE. W. A. Farmer, 141 
Grand Ave., Rochester 9, N.Y. 

Cameraman-Director: Currently em- 
ployed by internationally known pro- 
ducer, desires greater production oppor- 
tunities. Fully experienced 35- and 
16-mm, color, b & w; working knowl- 
edge editing, sound, and laboratory 
problems; administrative experience. 
Top references and record of experience 
available. Write P.O. Box 5402, Chi- 

Cameraman: Trained with practical 
experience in 16-mm and 35-mm equip- 
ment & technique with prominently 
successful men in the industry. Thor- 
oughly familiar with B & H Standard, 
Mitchell, Eyemo, & Filmo cameras, 
Moviolas, etc. Thorough knowledge 
& experience script-to-screen produc- 
tion technique: directing, editing, 
photography, film evaluation, produc- 
tion, treatments, shooting-scripts, small 
budgets, documentary & theatrical 
production. Go anywhere. Age 33. 
Top industry & character references 
furnished confidentially. Anxious for 
position where ability, sincere interest 
and creativeness offer opportunity. 
Active Member of SMPTE. Write 
Milton L. Kruger, R.F.D. 1, Ridge- 
wood, N.J. 


By action of the Board of Governors, October 4, 1931, this Honor Roll was 
eslablished for the purpose of perpetuating the names of distinguished pioneers 
who are now deceased. 

Louis Aime Augustin Le Prince 

William Friese-Greene 

Thomas Alva Edison 

George Eastman 

Frederic Eugene Ives 

Jean Acme Le Roy 

C. Francis Jenkins 

Eugene Augustin Lauste 

William Kennedy Laurie Dickson 

Edwin Stanton Porter 
Herman A. DeVry 

Robert W. Paul 
Frank H. Richardson 

Leon Gaumont 
Theodore W. Case 

Edward B. Craft 
Samuel L. Warner 

Louis Lumiere 

Thomas Armat 


Lee de Forest 

A. S. Howell 


Journal of the Society of 

Motion Picture and Television Engineers 


Television Cutting Techniques RUDY BRETZ 247 

Spontaneous Ignition of Decomposing Cellulose Nitrate Film 


Sensitometric Aspects of Background Process Photography 


A Motion Repeating System for Special Effect Photography 

0. L. DUPY 290 

Increased Noise Reduction by Delay Networks 


Miniature Condenser Microphone JOHN K. HILLIARD 303 

Supplementary Magnetic Facilities for Photographic Sound Systems . . 

G. R. CRANE, J. G. FRAYNE and E. W. TEMPLIN 315 

Sprocketless Synchronous Magnetic Tape R. H. RANGER 328 

A New //1 .5 Lens for Professional 16-Mm Projectors. . .W. E. SCHADE 337 

The Metal-Diazonium System for Photographic Reproductions 

R. J. H. ALINK, C. J. DIPPEL and K. J. KEUNING 345 

Proposed Bylaw Amendment 367 

16-Mm Sound Service Test Film 375 

Characteristics of Color Film Sound Tracks COMMITTEE REPORT 377 

67th Semiannual Convention 379 

Society Announcements 380 

Film Decomposition Tests 381 

Meetings of Other Societies 383 

Engineering Committees 384 


Introduction to Theoretical and Experimental Optics, by Joseph Valasek . . . 

Reviewed by Dr. K. Pestrecov 385 

Letter to the Editor By A. I. Mahan 387 

Section Meetings 387 

New Products 388 

Employment Service 390 


Chairman Editor Chairman 

Board of Editors Papers Committee 

Subscription to nonmembers, $12.50 per annum; to members, $6.25 per annum, included in 
their annual membership dues; single copies, $1.50. Order from the Society's General Office. 
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and 
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture Engineers. 
Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General and Editorial Office, 
342 Madison Ave., New York 17, N. Y. Entered as second-class matter January 15, 1930, 
at the Post Office at Easton, Pa., under the Act of March 3, 1879. 

Copyright, 1950, by the Society of Motion Picture Engineers, Inc. Permission to republiah 
material from the JOURNAL must be obtained in writing from the General Office of the Society. 
Copyright under International Copyright Convention and Pan-American Convention. The 
Society is not responsible for statements of authors or contributors. 

Society of 

Motion Picture and Television Engineers 

342 MADISON AVENUE, NEW YORK 17, N.Y., TEL. Mu 2-2185 

Earl I. Sponable 
460 W. 54 St. 
New York, 19, N.Y. 

Peter Mole 

941 N. Sycamore Ave. 
Hollywood 38, Calif. 



Loren L. Ryder 
5451 Marathon St. 
Hollywood 38, Calif. 

Clyde R. Keith 
120 Broadway 
New York 5,'N.Y. 

Robert M. Corbin 
343 State St. 
Rochester 4, N.Y. 

William C. Kunzmann 
Box 6087 
Cleveland 1, Ohio 

Fred T. Bowditch 
Box 6087 
Cleveland 1, Ohio 


Frank E. Cahill, Jr. 
321 W. 44 St. 
New York 18, N.Y. 

Ralph B. Austrian 
25 W. 54 St. 

New York 19, N.Y. 


Herbert Barnett 
Manville Lane 
Pleasantville, N.Y. 

Kenneth F. Morgan 
6601 Romaine St. 
Los Angeles 38, Calif. 

Norwood L. Simmons 
6706 Santa Monica Blvd. 
Hollywood 38, Calif. 


George W. Colburn 
164 N. Wacker Dr. 
Chicago 6, 111. 

Charles R. Daily 
113 N. Laurel Ave. 
Los Angeles 36, Calif. 

John P. Livadary 
4034 Cromwell Ave. 
Los Angeles, Calif. 

William B. Lodge 
485 Madison Ave. 
New York 22, N.Y. 

Edward Schmidt 

304 E. 44 St. 

New York 17, N.Y. 


Lorin D. Grignon 
1427 Warnall Ave. 
Los Angeles 24, Calif. 

Paul J. Larsen 
508 S. Tulane St. 
Albuquerque, N.M. 

William H. Rivers 
342 Madison Ave. 
New York 17, N.Y. 

Edward S. Seeley 
161 Sixth Ave. 
New York 13, N.Y. 

R. T. Van Niman 
4501 Washington St. 
Chicago 24, 111. 

Television Gutting Techniques 




Summary The techniques of motion picture cutting have in some measure 
found application in television cutting as well. Characteristics peculiar to 
the new field, however, have influenced the growth of new techniques. In 
this article the author outlines the basic principles of television cutting from 
the creative or production point of view. In a succeeding article he will 
discuss the equipment in use in television stations for the switching of 
picture signals. 

IN MOTION PICTURES, a change from one scene to another is usually 
accomplished by a cut. The editor determines the exact point at 
which the change should occur, trims off the unnecessary portions 
from the end of the first scene and the beginning of the second, and 
then joins the two together with a splice. This process is called 
"cutting." To "cut" a film does not mean simply to reduce it in 
length by taking out unwanted portions. It is the whole process of 
putting a film together, choosing which shot shall follow which, 
determining the precise length of each and the exact frame on which 
each cut shall be made. 

The cutter is the one who does the physical work of handling the 
film. The creative decisions are usually made by a "film editor," who 
may sometimes do his own cutting. Motion picture cutting is recog- 
nized as a vitally important aspect of the film art. It is a profession 
in itself. Some film makers have gone so far as to say that in the 
technique of cutting lies the entire process of creative film art. 

In television, a change in the scene is made with an electric switch, 
not with the scissors and splicer. Strictly speaking, we should call 
it "switching" instead of cutting. But the word "cut" has come to 
mean an instantaneous change of scene; and, since that is what we 
produce by pushing buttons, it is still correct to call the process cut- 
ting. Both terms will be used hereafter: "cutting" when the crea- 
tive aspects of the process are being considered, "switching" when the 
discussion is on the technical side. 

A CONTRIBUTION: Submitted January 10, 1950. This is part of a forthcoming 
book and is published by permission of McGraw-Hill Book Co., Inc. Critical dis- 
cussion of this material is particularly invited, either in the form of Letters to the 
Editor or by communication directly to the author at Croton-on-Hudson, N.Y. 

248 RUDY BRETZ March 

In end result, the process of cutting television is very similar to that 
of cutting motion pictures, but the method is entirely different. 
A film editor may take weeks to cut a show, deliberating over each 
decision. A television director has to do his job during the show 
itself. He must rely on fast thinking, quick reaction time, and thor- 
ough preparation. In some cases he relies on his script, or on his 
assistant, who follows the script. The sight of the television director 
excitedly calling shots ("Take Two! Take One!") has led some to 
think that shot calling is all there is to television directing. Some 
directors actually do forget to make full use of their cameras, because 
they are so occupied with just the problem of cutting. 

In creative production a good director plans his cuts, just as he 
plans his camera shots : on paper. If he has had enough experience, 
he can visualize the effect of a cut, even between two shots which also 
exist only in his mind, or as thumbnail sketches on the margin of the 
script. He will not limit himself to what he has planned, however, 
once he has produced the shots with cameras and watched the effect of 
the cut on the master monitor. He may change the camera shots to 
improve the cut, or he may change the timing of the cut itself. 

By far the greatest number of television shows, however, are pro- 
duced with little or no rehearsal; and the cutting cannot be planned 
in advance. Even if camera shots are set in rehearsal, the cutting is 
very rarely the same on the final air show. This is because the timing 
of the cut or the choice of camera depends mainly on the performers. 
If they deviate from the rehearsed routine, or go through actions 
which have not been rehearsed at all, the director must be on his toes 
to adapt his cutting to their performance. This is called "off-the- 
cuff" shooting, "ad lib" cutting, or "winging" the show. The method 
of control-room operation makes a lot of difference in whether this 
kind of cutting can be successful. 

1 . The Switcher System 

Under this system the director is in constant inter-phone connec- 
tion with the cameramen. He is responsible for the placement and 
use of the cameras, and is considered in direct charge of the camera- 
men. He "calls his shots" to an engineer, whose function it is to 
punch the buttons which switch from camera to camera. This engi- 
neer is sometimes called the "switcher," sometimes the "technical 
director" (T.D.). This is the most common method in use today. 


2. The Technical Director System 

Under this system the director has no verbal contact with the 
cameramen. He will call shots to a technical director who punches 
the buttons as in the preceding case. The technical director under 
this system, however, is in charge of the cameramen and the use of the 
cameras, as well as all other studio facilities. He works with the 
cameramen as a team and with the director as co-worker. Some- 
times the T.D. will take cuts himself especially if they have been 
established in rehearsal. Usually he will punch buttons only at the 
director's command. The director may talk to the studio during 
rehearsal over a loudspeaker, and he can talk to the floor manager 
over earphones. As far as I know this method is limited to the five 
NBC-owned stations across the country, to WABD in New York and 
WSB in Atlanta. It has been tried elsewhere and abandoned. 

Proponents of this system see in it an analogy to the method of 
Hollywood production, in which each film has two directors, one of 
whom is the Director of Photography and is in charge of all technical 
aspects of the production, while the other carries the title of Director 
and is responsible for the dramatic action. Some directors, who have 
also worked under the switcher system, prefer this because it makes 
their work lighter. Relieved from having to think about cameras 
and cutting, they are able to do a better job with the directing of the 

One director, who has directed programs on many stations, says 
that working under the NBC system is like having a twin directing 
the program with you. During rehearsal much time can be saved. 
When in the course of rehearsal a stopping point is reached, the direc- 
tor can go out and fix the actors' errors and come back to the control 
room to find all the camera changes made and everything ready to go 
again. Of course this is predicated on the ability of the T.D. He 
must have almost as good a background as the director himself. He 
must be primarily a showman, not an engineer. Where this method is 
in use, however, the job of T.D. is always an engineering job. The 
T.D. must be in charge of the camera and control-room crew, and for 
this reason must be a superior engineer. Men who can fulfil all these 
requirements are rare, or cannot be found for the salaries that are 

The T.D. system breaks down when an inexperienced man is on the 
job, or when an ad lib show is attempted. I have watched a green 
director, a green T.D. and green cameramen attempt to use this 

250 RUDY BRETZ March 

method, and the results were miserable. The director would give a 
camera order, and the T.D. would garble it a little in relaying it to the 
cameraman, since he had no clear idea of what was meant. Then the 
cameraman would do the wrong thing. "No! I didn't mean that!" 
the directorwould tell theT.D. "I meant so-and-so!" This was again 
relayed to the cameraman. This time the cameraman himself 
would make an error. When it was all finally straightened out and 
everyone knew what it was the director wanted, it would turn out to 
be something that couldn't be done anyway because of some technical 
factor which no one had anticipated. 

During an ad lib show, camera and cutting instructions must be 
given very rapidly and acted upon immediately or action may be lost. 
Sudden instructions to the cameramen cannot originate in the director 
in these cases, since, by the time they are relayed through the T.D., 
the moment has passed. To be sure, the T.D. himself may wing the 
show; but he is acting then in the capacity of director, which very 
few T.D.'s are able to or would be allowed to do. A good T.D. could 
assist the director by keeping one step ahead of him. He could so 
engineer the cameras that the director would have a variety of shots 
to choose from in calling takes and one camera at least would always 
have a good shot, well framed, from the proper angle to show the 
action, and ready at the right time. There is some question, however, 
as to whether this could not be done just as well or more easily by the 
director himself. The general opinion is that for ad lib shows the 
T.D. system does not work. Since most television stations must do 
the ad lib type of production (and practically all remote pickups fall 
into this category), most stations have decided against this method. 

3. The Director-Switching Method 

In this method, the director not only talks to the cameramen but 
operates the switching system as well. Some of the smaller stations 
have adopted this method purely for reasons of economy, and because 
there was no local union jurisdiction over the job. Others have 
chosen it because they believe it to be the best method of control- 
room operation. The exact moment that a switch is made is very 
important, and a split second's delay can often mean the difference 
between a good and a bad cut. In baseball pickups, a bad cut is 
much worse than that: it is a lost play. A majority of the stations 
have adopted the director-switching system at least for baseball and 
other sports remotes. A few have carried the policy over into studio 


operation. WCAU-TV and WPTZ in Philadelphia are good ex- 
amples. Making a comparison with the T.D. system, they say at 
WPTZ, "It is easier to train a director to punch a few buttons than it 
is to teach an engineer showmanship." 

This method of operation works out fairly well in small stations and 
on programs which are not too complicated. In the production of 
dramatic shows, however, it is advantageous to the director to be 
able to dispose of the burden of handling the dials and switches. 


The principles of cutting that I shall outline here are the same that 
have governed film editing in all but the most esoteric types of films 
or film sequences. These are the techniques that achieve a "smooth 
cut" and assure a visual continuity. Good cutting, under this 
criterion, is unnoticeable. Two shots are joined in such a way that 
the audience is completely satisfied by the result and its attention can 
properly remain with the action, rather than being distracted by the 
method of production. Subject is more important than form in this 
type of production. 

Basically, the responsibility of the television director is to satisfy 
the viewer. When something is going on and the director is making- 
pictures of it, he must show it properly, or the viewer will be dissatis- 
fied. He must show the viewer what he wants to see. Forty years 
of an ever-improving motion picture art has educated him to expect a 
lot. He wants close-ups on essential action and he wants them 
quickly, just as he is used to getting them in the films. He wants to 
look around and know where he is and he expects good orientation. 
Above all, he doesn't want to miss anything that happens and he 
doesn't want to be confused. 

The director must show the viewer what he wants to see, when he 
wants to see it, and cause him no confusion in the process. If the 
television director can achieve even so much as this, he will be a good 
director. Fancy angles, subjective camera and montage cutting also 
have their place, but they can never substitute for the basic require- 

Showing the viewer what he wants to see might better be classified 
under camera handling than under cutting. But showing it to him 
when he wants to see it is definitely a principle of cutting. The timing 
of the cut is perhaps the most important single thing about it, and the 

252 RUDY BRETZ March 

one thing that is most likely to be a little off in television. Film edit- 
ing procedures take account of the timing factor and usually include a 
"rough cut" stage where all the scenes are overlong. When the film 
is projected and studied in this condition, cuts can be more exactly 
timed. In television such a procedure is impossible. A long shot 
and a medium shot which are to be cut together exist side by side on 
two monitors. A switch between them is instantaneous and irrevo- 

Within a sequence of shots taking place in one scene, the actual 
length of an action cannot be changed. An actor crosses the room, 
and we cut to a close-up when he reaches the other side. In film 
editing there is a possibility of shortening the actual time of the cross. 
If the actor walks out of the frame on the first shot and into the frame 
on the second, the entire intervening time can be eliminated. This is 
known as "creating filmic time." In television we can condense 
time in this way only between sequences or by special devices, but 
not in the regular run of shooting. We are limited to actual time, 
since television is basically an art of actuality. The choice, by the 
director, of the best segment of this actuality, at the best time, is the 
process of artistic selection, which is a good part of what might be 
called the creative art of television. 

Dependence on actual time, however, simplifies television cutting 
as compared to cutting film; there is no need to worry about matching 
of action. In film shooting, it is always a constant worry to be sure 
that the action is the same when you shoot a close-up as it was when 
you shot the medium-shot just before. If an actor sits down, he 
must do it at the same speed each time, hold the chair with the same 
hand, cross his legs in the same direction, etc., or joining the two 
pieces of film will be a great problem. Then, in film cutting, even 
when you have two shots with identical action in them, there is the 
problem of cutting between them, so that the action, which begins in 
one shot and ends in the other, will be smooth and continuous. There 
must not be any overlapping or anything missing. This also we are 
spared in television. 

Cutting on Action 

The principle of cutting on action is just as important in television 
as it is in films. There is nothing that will so disguise the fact that a 
cut has been made as a strong and positive action to carry across 


from one shot to the next. The cut should be made during the action 
itself, not just before it, and not after it. As a director you must 
watch your monitors very closely, have the second camera ready, call 
the first half of your order, "Take ...," and just at the moment of 
the move, call the camera number. If it is a short movement, like a 
dancer's leap, for instance, and there is any delay at all, the cut will 
come after the leap is over. 

It is very important for the switcher or technical director who 
punches buttons to know ahead of time which camera is to be taken 
next. His reaction time is faster then. He has only to punch the 
button. He doesn't have to decide first which button to punch. 
KLAC-TV in Hollywood has established a rule requiring directors to 
call the camera number first, before they call the take. If they say, 
"Two. Take Two," it gives the technical director time to get ready, 
and a well-timed cut results. In spontaneous cutting, however, it is 
not always possible to anticipate sufficiently ahead of time. Many of 
the directors' orders at this studio sound more like "Twotaketwo," 
which is no great advantage, and just gives the director more to say. 

The director's best method, I believe, is to give a "ready" cue when- 
ever possible. This is appreciated also by the cameraman who is not 
so likely to be caught changing lenses or adjusting focus at the moment 
of the cut. "Ready Two" can be said almost automatically while you 
are watching the action on camera Two. Where the cut must be 
precise, a further "Take ..." will keep the technical director poised. 
Then the number can be thrown very quickly, accompanied if desired 
by a hand signal, and the reaction is almost as fast as though the direc- 
tor himself pushed the button. 

The moment of action is such a natural place to cut that it is pos- 
sible to violate the other principles of good cutting and get away with 
it, as long as the cut is made on action. For instance, one principle, 
which I shall describe later, is concerned with keeping left and right 
straight in the viewer's mind. Whoever is looking or moving left in 
one shot should be looking or moving left in the next, if the action is 
to match at all. However, Don Hallman at WXYZ-TV covers 
wrestling with two cameras, which are on opposite sides of the ring. 
The purpose of this setup is that one camera can see what is hidden 
from the other. He has great problems in cutting, however. When- 
ever he switches cameras, the contestants change places on the screen. 
Hallman's method is to wait for a positive action before making his 




switch. A definite action (and there are many such movements in 
wrestling) will leave no doubt or confusion in the mind of the audience 
as to which person is which. 

Cutting on action is possible, however, only if the attention of the 
audience is definitely centered on the action through which you in- 
tend to bridge the two shots. If you start with a long shot which in- 
cludes several actions, the audience is as likely to be watching one as 
another. If the shot, let us say, shows four football players warming 
up before a game by kicking punts, and the director should wish to 
cut to a close-up of one of them, he cannot do this on the action of the 
kick. There is a three-to-one chance that the viewer may not be 
watching the same player that the director is looking at. If the 
viewer's attention happens to be on the wrong player, the new shot 
will suddenly be upon him in the midst of an action. It would be 
best in this case to cut before the kick, so that the complete action 
would be included in the close-up shot, just as one would do in cutting 
to another scene, when matching of action is unnecessary. 

Figure 1 

Cutting on Reaction 

One of the most powerful motivations for a cut is to have someone 
in the picture look outside the frame. Immediately the audience 
wants to see what he is looking at. A shot of another subject any- 
thing under the sun is accepted, at least momentarily, by the eager 
viewer. The viewer has been shown what he wants to see. More 
than that, the director has contrived, by the device of having some- 
one look, to make the viewer want to see the thing the director is going 


to show him next. He can cut to almost anything at all and make a 
good cut. For a joke, he can even put in something ridiculous. If 
the shot is entirely impossible and incongruous, the viewer will laugh 
at himself for having accepted it, but only after a smooth and natural 
cut has been foisted on him. 

This same general principle applies in the case of an actor pointing a 
camera or a gun. The audience wants to see what the actor is aiming 

Such a device can be used to motivate a cut to a very big close-up, 
from a medium- or long-shot. Usually it is best not to make the 
change of shot too extreme. Long-shot to medium-shot, medium-shot 
to close-up, close-up to big close-up is the normal progression. But 
if a person picks up a picture or a letter or obviously concentrates his 
attention on a very small portion of the scene in front of him, a cut to 
that small area is perfectly motivated. 

Cutting is often motivated when the audience wants to see some- 
one's reaction to what has just been said or done. In the audience 
participation show, a contestant is blindfolded and put through some 
silly stunt. A shot of his wife's reaction to his asininity is completely 
motivated (provided she is reacting at all). In the dramatic show, 
when two people are in conversation, it is often desirable to cut to a 
close-up of the person who is not talking, in order to watch his reac- 
tion to what is being said. Standard procedure under other circum- 
stances, when covering a conversation, is to show the person who is 
talking, at all times, and to cut back and forth on every speech. 
You have to show the viewer what he wants to see, and he is inter- 
ested primarily in the source of sound. Noises off screen, the opening 
of a door, or a voice, will make the audience want to see where the 
sound is coming from. 

The Cut-in or Intercut Shot 

Sometimes during the course of an event or performance it may 
enrich the show to cut away from the event and interpose a shot of 
something else. A spectator reacting to a sports event, or not react- 
ing, as the case may be, is a good example. The newsreels frequently 
intercut a close-up of screaming fans just after a good play or show a 
long-shot of the stands if the preceding action has not been too excit- 
ing. But in films this serves a practical purpose, which does not hold 
true in television. 

256 RUDY BRETZ March 

An intercut shot in a film can be used to separate two shots that do 
not match and cannot be cut together. A smooth continuity results. 
Moreover, the film audience does not mind being taken away momen- 
tarily from the primary action, because it knows that the film editor 
has included the rest of the scene in the reel, and that it is not going 
to miss anything. The television viewer, on the other hand, feels no 
such certainty. He is afraid, while he is watching the frenzied fans 
jumping in the stands, that he is missing something going on down on 
the field. The same is true of audience participation shows, comedy 
shows, variety shows and many other kinds of spontaneous programs. 
If intercut shots are to be used at all, they must be used at a time 
when nothing of importance is likely to happen elsewhere, or they will 
not be what the viewer wants to see. 

A dramatic show is much more akin to a film in this regard. The 
thing is built and presented as a whole. It is not segments of reality, 
and the audience has no fear of missing anything. 

All of the cinematic techniques are applicable to television dramatic 
shows, if they can be physically accomplished at all. Intercutting of 
extraneous shots for purposes of contrast, irony, flashback, etc., can 
be accomplished as well in television as in the film medium. These 
techniques are analyzed in such books as Rudolf Arnheim's Film, 
Raymond Spottiswoode's A Grammar of the Film and Ernest 
Lindgren's The Art of the Film. 

Length of Shot 

Among people who don't know too much about the art of cutting, 
there is prevalent a misapprehension that a lot of cutting increases 
the tempo of a production. This is carried so far that statements 
like this have been heard: "There should be a cut at least every 20 
seconds in order to keep audience interest." This is by no means 
always true. A shot should be as long as the proverbial piece of 
string. Tempo is not controlled only by the rate of cutting, except 
perhaps in newsreels or documentary films. For instance, in the news- 
reels or in the "March of Time," films which are built from a series of 
more or less disconnected shots, scenes of three or four seconds in 
length are standard. A shot containing action should be continued 
as long as necessary to complete the action, or cut as short as possible, 
one might rather say, without losing any of the action. "The True 
Glory," a war film made from thousands of stock shots taken by a 
great number of Signal Corps cameramen, was cut to a very rapid 


pace. The shots were, straight through the film, less than two seconds 
in length. This resulted in a picture which was very hard to watch. 

Cutting, in this type of film, does control tempo, because it controls 
the speed at which the film progresses from subject to subject. Tele- 
vision production, however, is not like this at all. Cutting, in tele- 
vision, is usually not from one subject to another, but from long-shot 
to medium-shot to close-up, etc., all on the same subject. The speed 
at which the actors in the play progress from action to action is what 
determines tempo. All the frantic cutting in the world can't speed 
up the show. 

I remember once, as a young director, being given a sports inter- 
view show to direct. The master of ceremonies and the interviewee 
sat together on a sofa and talked. Here was my great opportunity to 
do a real production. I dollied the cameras in and out. I cut from 
one angle to another: big close-ups, high two-shots, timing each cut 
accurately with the phrasing of the conversation, building up what I 
thought must be a terrific pace. But I succeeded only in sweating 
up the camera crew and making it hard for the audience to relax and 
pay attention to the interview. The show remained nothing more 
than two people sitting on a sofa and talking. 

What is the absolute minimum length of shot? is a question that is 
sometimes asked. Four frames of film are enough to give the film 
audience a glimpse of a subject; there are shots of this length in 
some of the montage sequences in Hollywood films. It is amazing 
how quickly the eye can take in a picture. Back in the Keystone 
Comedy days they found they could get a better laugh if they cut 
in close-ups of action. If the action was only a pie hitting a face, 
or a quick change of expression on a comedian's countenance, it was 
included in a few frames of film, and there was no need to run the 
shot even as long as one second to show what had happened. 

Fast cutting is very rarely desirable in television, and hardly ever 
possible. For one thing, you run out of shots too fast. You have 
only two or three cameras, and you have to allow time between takes 
for the cameramen to line up new shots or you will be repeating your- 

Furthermore, television doesn't call for the pace of the motion pic- 
ture. The two media are different in this respect. Since the audience 
is watching reality in the case of television, it is content to let events 
take their own natural time. It is hard to imagine a group of people 
sitting through a two and a half hour film of a football game in the 

258 RUDY BRETZ March 

way that they will watch a full-game telecast. As movie-goers, they 
are used to a condensation of time and a tempo of production which 
is purely filmic. 

Cutting vs. Camera Movement 

Since the film is a. construction made out of shots spliced together, 
it is natural that the cut would be accepted as the normal thing. 
Television, on the other hand, is an electrical pickup of reality. The 
television camera makes pictures continuously, as long as it is turned 
on and the beam control is up. Theoretically, the long continuous 
shot is more natural to the television medium. Furthermore, a tele- 
vision show is much less a construction than it is reality itself. The 
television viewer feels toward the receiver as though it were a kind of 
window through which he can see distant events. The more real 
this view, the stronger his feeling. These theoretical considerations 
seem to indicate that cutting is something to be avoided in television. 

Camera movement is, of course, the alternative. Instead of cutting 
to a closer shot, why not dolly in? Is it not better to use a Zoomar 
lens on a baseball game than to cut from camera to camera? By and 
large, this is true. Tony Bunzman, who directed at NBC before the 
war, did some very fine dramatic shows with only one camera. 
There are times, however, when the advantage of a cut over camera 
movement is very great. Several such advantages can be listed: 

1. One practical consideration is time. It takes time to dolly in to a 
close-up, and back again. The action must slow down and wait for 
the camera. It is much better to cut to a close-up and cut away from 
it again in a matter of seconds, and get along with the show. 

2. Rehearsal tune is also involved. A dramatic show can be 
planned with a great deal of actor and camera movement, or it can 
be done by cutting between a series of static shots. A lot more 
rehearsal time is needed to produce a show with camera movement. 
The movement, the positions of the actors and the co-ordination of 
the cameras, all have to be carefully worked out and well rehearsed if 
they are to work at all. Simple cutting from one shot to another can 
be worked out in much less time. 

3. Reaction shots, also, often cannot be made by panning or dolly- 
ing the camera. When the viewer sees someone look at something, 
he wants to see immediately what the actor is looking at. The viewer 
is seldom content to wait through a long pan shot across unimportant 
background to see it. 




4. Another consideration is the dramatic value of the cut itself. 
The sudden appearance of a new picture on the screen can be put to 
good use sometimes, entirely for its shock value. This is particularly 
possible if cuts have been used sparingly in the sequence just before. 
A sudden dramatic moment can be enhanced by sharp cuts. A cut is 
useful for punctuation, something that is more difficult to accomplish 
with camera movement. 

Matching the Center of Interest 

When a cut is made, the eye must quickly adjust itself to a new 
composition. The easier the readjustment, the less noticeable and 
the smoother is the cut. One factor which determines this smooth- 
ness is the relative position in the frame of the center of interest in 
each case. In the first shot in Fig. 2 the center of interest is the 

Figure 2 

woman talking, in the lower right corner of the screen. If a cut is 
made to a close-up of the same subject, but composed so that the 
woman is then to the left of the screen, a readjustment of the eye is 
necessary. The eye remains focused on the lower right corner after 
the cut, with nothing particular to look at. It must find the center 
of interest again in its new position on the screen. 

This is really a minor point in television, since there are usually so 
many other, more important things unaccomplished. Where greater 
perfection is desired, however, it is something to consider. An ex- 

260 RUDY BRETZ March 

ample is in the pickup of sports. The smoothest possible cutting is 
necessary in covering sports because even a momentary confusion is 
enough to cause the viewer to lose track of a play, particularly in a 
fast game. It is about all you can hope for if each camera merely 
centers on the action. If each camera centers the ball, at the moment 
of the cut the ball will remain in the same place on the screen. 

Cutting to the Audio 

Sometimes, when all else fails, the phrasing of sounds will provide a 
natural place to cut. For instance, all else being equal, it is smoother 
to cut at the end of a sentence than in the middle of one. The phras- 
ing of music forms a very compelling pattern for cutting, especially 
when the music plays an important role. The music is reinforced if 
the visual change follows the changes of tone color, chorus and verse. 
Clark Jones, one of the best directors in this field, is known for his 
accurate cutting in musical shows. His shots may be eight bars, 
four bars, sometimes only two bars in length, depending on the change 
in tone color of the music, and the cut is always made precisely on the 
beat, showing the new musician just as he begins to play. When 
tone-color changes every two bars, so do the cameras; when a vocalist 
sings 32 bars without a break, the camera rests quietly with the sub- 

What Not to Do in Cutting 

To the tyro director who has had little experience in the film or 
television media, I would direct the following words of caution : 

1. Don't cut too much. You are likely to make the show harder to 
watch, or to irritate your audience. If you have time to work it out, 
camera movement or actor movement is smoother. Use the cut for 
dramatic punctuation or to get a shot on the air quickly. Don't go 
crazy with "brilliant" cutting unless you know what you are doing. 

2. Don't cut blindly. Some directors seem to close their eyes and 
say, "Take Two." They are afraid to take their eyes off the master 
monitor. Keep looking back and forth, try to watch all the monitors 
at once. And look at the camera monitor before you call the take. 
You'll get fewer lens changes and out-of-focus shots on the air, if you 

3. Don't cut from one shot to another one just like it. A cut from a 
two-shot to a two-shot adds nothing. To be sure, the second one is 


from a different angle, but does it show the audience anything more? 

Never let two cameras give you the same shot. Be sure that one 
of them changes lenses or repositions, otherwise it might as well not 
be out there. Sometimes a director gets stuck and has to cut be- 
tween two identical shots, bad as such a cut may be. He may find, 
for example, that he has to release the camera which is on the air for 
some sudden need. He must cut to another camera. His only possi- 
bility is an identical shot on another camera, and he hasn't time to 
change it to something different. The viewer doesn't hear his 
excuses, however; all he sees is a bad cut. 

The worst example I can remember of cutting between similar 
shots occurred on a musical program. In this case it was close-ups 
of a singer. There were three cameras: one presenting a full face of 
the girl, one a left profile the same size, and the third a right profile, 
also the same size. During the song the director changed shots often. 
There might have been a small reason for this because of the variety 
obtained by the change of angle, except that the performer was too 
smart for them. She was on to television. As soon as the camera 
changed, she spotted the red lights out of the corner of her eye and 
turned to face the new camera. Thus wherever the director shot 
from, he always got a full-face close-up. The singer continued to 
chase the cameras around for the rest of the number. 

4. Don't cut to an extremely different angle. What constitutes too 
great a change in angle is hard to define. The main thing is to be 
sure that the subject is immediately recognizable and doesn't look 
like something else in the new shot. A shot from an angle that is so 
far different that an altogether different background is seen behind 
the subject will cause confusion. Sometimes a change of angle will 
make a great difference in the lighting. What is contrasty side-light 
from one angle may be flat front-light from another. The difference 
may be so great as to make the subject unrecognizable, at least for a 
moment. I have seen a profile shot of a news commentator cut in 
suddenly after we had been watching a full-face shot, with the result 
that the profile shot looked momentarily like someone else sitting 
across the room watching. The audience has no way of knowing that 
the new shot is taken from a different camera angle until they have 
seen the shot and comprehended it. The natural assumption is that 
it is taken from the same place, but looking in a different direction, 
just as an observer on the scene would turn and take in another view. 




The problem of camera placement in the pickup of baseball and other 
sporting events is directly tied in with this. When an action takes 
place between two cameras, so that they see the subject from opposite 
sides, the direction of the action will be completely reversed when a 
cut is made. Nothing can be so utterly confusing. It is usually best 
to keep at least the main two or three cameras as close together as 
possible on most sports pickups. 

Figure 3 

5. Don't cut on a pan. Don't cut from a camera that is panning to 
one that is static, or from a static camera to one that is panning. Try 
it once and you will see what I mean; it just can't be done. You can 
cut from pan to pan very nicely, providing they are both going in the 
same direction and at the same speed. When two cameras are follow- 
ing the same action, for example, the cut is usually quite smooth. At 
a football game, if both cameras center the ball as they follow an end 
run, it is perfectly possible to cut from medium-shot to close-up with- 
out the slightest confusion, since both cameras will be panning, and 
in the same direction. 

6. Don't spoil exits. As far as the audience is concerned, once an 
actor has left the frame of the picture, he has made his exit. Don't 
rediscover him and exit him again in the background of the next shot. 


The Dissolve 

The dissolve, lap dissolve, or cross-fade is the method of transition 
in which the first picture becomes steadily weaker while the second 
becomes stronger on the screen. It is really a fade-out simultaneous 
with a fade-in. If you stop the process at the mid-point you have a 
superimposure : each picture at half strength. 

In films, the dissolve has been used with a certain connotation. By 
unwritten agreement, all film makers found it best to reserve the dis- 
solve for a particular purpose. (The expense of having dissolves 
made assisted toward this end.) The meaning of a dissolve in films is 
transition through space and time. You dissolve between sequences: 
to a later time or to another place, or both. Dissolves within a 
sequence where no lapse of time is intended are rarely used in films, 
unless it may be by an amateur, fascinated with the dissolve mecha- 
nism on his camera, who dissolves between the long-shot of a garden 
and the close-up of a flower. 

In television, unfortunately, a dissolve is no harder to do than a 
cut, and costs no more. Instead of taking another day to accomplish 
it, as we do in films (taking out the proper negatives, sending them 
out for opticals, then splicing back the dupe negatives into the reel), in 
television you simply turn a dial, or move a handle, as the case may 
be. With the Dumont equipment, as I shall explain in a later chap- 
ter, you punch the same button you ordinarily punch for a cut, and a 
dissolve is automatically effected. It is much too easy. 

The result is a very free use of dissolves, often in places where a cut 
is really preferable. Dissolving within a sequence is often done on 
television, particularly in musical or variety shows. In such produc- 
tions the director usually feels less confined to the motion picture 
tradition than he does when he is producing a dramatic sho\v. The 
television dramatic show seems to follow the motion picture technique 
as closely as possible. The result is that in television we have a new 
connotation (or a loss of any connotation at all) for the dissolve. 

Don't go dissolve-happy. Don't use a dissolve when a cut would 
be better. Don't use a dissolve when camera movement would be 

The Fade-out and Fade-in 

The motion picture connotation of the fade-out has been retained 
in television. The fade-out is sometimes called the "dissolve to 

264 RUDY BRETZ March 

black" in television, because of the electronic means of effecting it 
with the dissolve control. It has a connotation of finality; it indi- 
cates an end to something. It may be used in place of a dissolve to 
link sequences together, where there is a greater change in time or 
place than is usually indicated by a dissolve. A dissolve retains con- 
tinuity; a fade-out-fade-in breaks the sequence. It is not good, for 
instance, to fade in and fade out a series of titles, since the audience 
will think with each fade-out that they have seen the last of the series. 
Similarly, a play divided into a number of short scenes with fades be- 
tween, will lack the unity it might have had if other devices were used. 

The most dangerous aspect of using fade-outs in television is that 
the length of blank screen between fade-out and fade-in may be too 
long. Audience interest drops very rapidly when there is nothing on 
the screen. There may be exceptions to this rule : certainly when sound 
or music carries through, the wait is not so bad. However, the timing 
of the fade-out, the blank screen, and the fade-in should always be 
carefully planned and rehearsed. The shortest possible blank screen 
is particularly important when the fade-out comes at the end of a 
program or spot announcement. Every second's delay in bringing 
up the next picture means an increasing loss of audience. Dial turn- 
ing on television is a much greater problem than radio ever had to 
contend with. Directors and technical directors have sometimes 
failed to look past the conclusion of one program and plan the initia- 
tion of the next. They wait until the final fade-out, heave a big sigh 
and then start hunting frantically through their schedule and routine 
sheets for the next move, while the screen remains patiently blank. 
This problem became so acute during the early months of operation 
at WPIX that we had to enforce a rule against any fade-outs at all. 
All transitions were to be made by dissolves. 

There are ways of doing fade-outs in a program other than by 
simply turning down the gain, or dissolving to black. Fade-outs can 
be done with the lights on stage: the old familiar black-out. This is 
easier if the number happens to be lighted by one spot light, which is 
all that needs to be cut. The light can be moved off the subject, or 
the subject can move out of the beam of the light. If the motivation 
is very carefully planned, the camera can be panned off into a dark 
area, or dollied behind a dark object. This device was used by 
Olivier in "Hamlet." Another variation on this is to have an actor 
walk directly into the camera until he blocks out all the light. Which- 


ever of these devices is used to effect the fade, it is usually good to use 
the same thing in reverse for the fade-in of the following scene. 

An unusual use of the fade-out-fade-in was developed by Bert 
Gold at WICU when that station had only one studio camera. In 
order to change lenses, or to go from one stage to the next, the picture 
had to be quickly faded, the change made, and the picture faded in 
again. This procedure called for close co-ordination with the camera- 
man, but was a good adaptation to the situation. Ideally, assuming 
complete freedom of camera movement and an expert cameraman, it 
should be possible to cover everything with one lens (say a 90-mm) 
and keep the camera on continuously. The need to fade the camera 
out and avoid showing lens changes or portions of the studio beyond 
the actual sets is an inheritance from the production techniques of 
other media. In television we don't have to shun the backstage view, 
except when illusion and mood would be broken. The more reality, 
the more immediacy, the better the television in most cases. 

Other Transitional Devices 

1. The white fade. To the best of my knowledge this was first tried 
by Carl Beier and Bernie Brink at CBS-TV in 1941. It consists in 
fading out to white instead of to black. Naturally when it was first 
discovered, everyone went white-fade-happy; and for a while it was 
used more than anything else, at least by the directors at CBS. 
NBC and Dumont, who were also on the air at that time, did not seem 
to think it was that good. 

Since the screen was constantly alight, it had less of a feeling of 
finality about it than a fade to black. The first time one saw it, 
however, one tended to feel that something had gone wrong and 
washed out the picture. We found good use for it between a series of 
titles (black letters on a white ground) so that only one camera and 
easel were necessary. It was a "fade white, pull the card, fade back 
in" routine. I haven't seen it used in recent years. 

2. The de-focus transition. This requires a little co-ordination with 
the cameraman. Ready the next camera you are going to use by 
having the cameraman put it out of focus. Then at the right moment, 
crank the first camera out of focus, cut or dissolve, and bring the 
second picture up sharp. It requires a series of cues such as this, 
"Ready Two out-of-focus. One, de-focus. Dissolve to Two. Into 
focus. Two." It is assumed that the cameraman on camera One will 

266 RUDY BRETZ March 

re-focus his camera again without being told, as soon as the effect is 
over. This device is particularly useful where a character in the play 
passes out or falls asleep just before the transition. The de-focus 
effect may also be combined with the fade-out for a greater transition 
through space or time. De-focus the camera, fade to black, then 
fade in to the next camera and focus up. 

3. Visual confusion or interference. Almost anything which fuzzes 
up or confuses the picture can be used very nicely as an alternative to 
the dissolve. It is best, of course, if the first appearance of the effect 
is motivated in some way. Whatever is at hand will do: people 
walking in front of the camera, fire, smoke, water, etc., in front of the 
lens. Distortion glass or crinkled cellophane might be used in front 
of the lens, but these special transitions are rarely utilized because 
they take time to rehearse and execute properly. Remember that 
the transition device may arise directly from the special effects within 
the show. Anything which confuses the picture will find its use some- 
where as a transitional effect. 

4- Fade-cut and cut-fade. These are additional combinations of the 
foregoing transitions, which have been used on television and should 
be included in this listing. It is sometimes desirable to fade out one 
scene slowly, and then jump right into another scene, using the shock 
value of the cut. Or a cut to black is sometimes indicated, where you 
want a sudden ending to a scene, followed by a fade-in to the next. A 
rather particular case of this can be taken from Ted Mill's "Garroway 
at Large" show. At the end of his program the casual Garroway 
picked up a camera cable that was lying across the studio floor and 
explained that it was the coaxial cable carrying the program across the 
country. "It's about time to end the program now," he said glancing 
at his watch and picking up a small hatchet. "Now this might hurt 
a little," he said. He brought the hatchet down and the picture went 
black. Then came a fade-in on closing titles. In a moment, how- 
ever, Garroway was back with his closing gag, holding the cable with a 
big bandage around it: "This program comes to you from Chicago, 
where we have very good connections." 

5. The wipe. This is a cinematic technique in which the new picture 
starts as a small area and grows until it covers the entire screen. 
Filmic wipes can be made in a great variety of patterns, and new ones 
can easily be devised to suit special purposes. The simplest is the 
plain horizontal wipe, which is the only one that has been done elec- 


Ironically in television. Several types of simple wipes can be done 
optically by shutters and masks in front of the camera. These will 
be discussed under special effects in another chapter in this book. A 
wipe between titles, for instance, is not to be confused with a pull-off 
or slide-through, which is a bodily movement of the title in front of 
the camera. In the true wipe, two cameras are used (or two stages 
of an effects machine) and everything is stationary. The only thing 
that moves on the screen is the line of demarcation between the two 
pictures as the area of one grows larger and the other smaller. 

No standard connotation was ever found for the wipe in motion 
pictures. It is very rarely used in dramatic films, finding its greatest 
value as a decorative device in industrial films, film commercials and 
the like, which depend on surface devices for their visual interest. 

Methods of doing an electronic wipe have been in the design stage 
for years. They were first put into regular operation by NBC in 
1949. The first use of the electronic wipe was to produce a split 
screen: with two ends of a telephone conversation shown at once. 
What was even more remarkable, an interviewer in New York was 
shown on the same screen as an interviewee in Washington. This 
will receive more complete treatment under electronic special effects 
in another chapter. All the switching systems at NBC-owned sta- 
tions (their own design) now are equipped with a dial control for mak- 
ing a horizontal wipe. A wipe is sometimes used to lead into a split- 
screen effect. A man starts a telephone conversation. On the op- 
posite side of the screen a wipe begins, then proceeds far enough to 
reveal the person at the other end of the line. When the call is over, 
the second picture wipes back out again. 

(A chapter on Switching is scheduled for next month.) 

Spontaneous Ignition of 
Decomposing Cellulose Nitrate Film 




Summary Cellulose nitrate motion picture film in the advanced stages 
of decomposition is liable to ignite spontaneously. The danger of such 
ignition is reduced by inspecting stored film stocks and removing and 
destroying all decomposing film. 

DURING THE ABNORMALLY HOT SUMMER OF 1949, numerous fires in- 
volving cellulose nitrate motion picture films were reported in 
New York City and adjacent areas. These fires occurred in process- 
ing and reclamation plants and in standard film storage facilities. 
Losses resulting from these fires, other than those of real property, 
were severe, for the majority of the films destroyed were original nega- 
tives and master copies, some dating back to the "silent era." Due 
to the fact that these fires occurred either after working hours or on 
week ends, no casualties to personnel resulted, although some firemen 
were treated for smoke inhalation. 

National Archives and Bureau of Standards engineers who investi- 
gated the fires could find no evidence that they were due to the negli- 
gence of personnel or the careless use of cigarettes or matches, but 
rather they appeared to have originated in the spontaneous ignition of 
deteriorated nitrate film. The summer of 1949 was one of the hottest 
recorded in the New York area, with the temperature reaching a maxi- 
mum of 98 F. The mean high temperature for the months of June 
and July, 1949, was 83.1 F, as compared to a normal for the period of 
79.3 F. The rainfall for the entire month of June was only 0.16 in. in 
contrast to the normal rainfall of 3.33 in. for the same period. The 
lack of rainfall and the unusually high temperatures of the period 
seemed to create ideal conditions for the development of spontaneous 
ignition in film stores. 

A CONTRIBUTION: Submitted February 9, 1950. 




Prior to the investigation of these fires it was generally believed that 
nitrate film did not ignite spontaneously at temperatures likely to be 
encountered in a normal film vault. At the request of the National 
Archives, the Fire Protection Section of the National Bureau of 
Standards instituted an investigation to determine the possibility of 
spontaneous ignition as an inherent hazard in decomposing nitrate 
film. Samples in various stages of decomposition were supplied by the 
National Archives for the purpose of simulating conditions which may 
have prevailed at the fire locations. These samples were stored in a 
special chamber, the temperature of which was controlled and re- 

Fig. 1. Container in which reel of film was placed, 
showing thermocouple wires. 

corded. The films were packed in individual cans with each wrapped 
in mineral wool to retain the heat of the exothermic decomposition 
reaction. The ambient temperature in the chamber was initially 95 F 
and, at intervals, was increased by small increments. After 17 days of 
this treatment one 1,000-ft reel of film, initially in an advanced stage 
of deterioration, ignited; the ambient temperature in the chamber 
at the tune was 106 F. Subsequently, with the ambient temperature 
at 120 F, another roll of film ignited. 

Tests made by the National Bureau of Standards have not yet been 
completed, but so far, some important conclusions can be drawn. 
Self-ignition temperature, which is dependent upon a number of fac- 
tors, was not the same for any two samples. The lowest temperature 




leading to ignition was 106 F. Because the number of samples investi- 
gated was small, it is doubtful whether this is the lowest temperature 
at which a reel of film can self-ignite. One reassuring aspect of the re- 
sults of the tests to date is that no film in good condition has self-ig- 

Fig. 2. Reel of film which had spontaneously ignited 
during a test at the National Bureau of Standards. 


At the moment, no one can predict what the weather conditions will 
be during the coming summer and it is quite possible that other re- 
gions may be confronted with abnormally high temperatures such as 
prevailed in the Atlantic Coastal Region during last year. This possi- 
bility offers the chance that there will be a recurrence of regrettable 
film fires. The hazard should not be underestimated for, even with 
abundant water supplies, cellulose nitrate fires are difficult to combat. 
Nitrate base film contains oxygen in chemical combination and does 
not need additional oxygen to sustain combustion. Furthermore, the 
fumes given off by its combustion are highly toxic and seriously 
hamper the fire fighters. They contain oxides of nitrogen which, if 




inhaled, can be fatal. Shortage of municipal water supplies in many 
areas presents an acute control problem definitely requiring the con- 
stant maintenance of every safeguard. 


The results obtained in the Bureau of Standards tests indicate that 
good film does not self-ignite at ordinary storage temperatures. 
Therefore, the logical approach is to remove from storage all film show- 
ing signs of deterioration. Such film can readily be found by regu- 









I. M|. M|.M I 


Fig. 3. Temperature during the critical period 
in which a film spontaneously ignited. 

larly scheduled inspection of stored film stocks. Inspecting personnel 
should be trained to recognize decomposing film by appearance, with 
its condition classified according to the following categories. In the 
first stage of deterioration the photographic portion usually shows an 
amber discoloration with fading of the picture image. In the second 
stage, the emulsion becomes adhesive and the film convolutions tend 
to stick together during unrolling. Rolls of third-stage film have an- 
nular portions which are soft, contain gas bubbles, and emit a noxious 
odor easily recognizable. In the fourth stage of deterioration, the en- 




tire film is soft, its convolutions welded into a single mass and fre- 
quently its surface is covered with a viscous froth. A strong noxious 
odor is given off, unmistakable to inspection personnel when once 
identified. In the fifth and final stage, the film mass degenerates par- 
tially or entirely into a brownish acrid powder. 

Deteriorated film in the first and second stages is photographically 
reproducible. If the matter recorded is important, the film can be 
copied and the original disposed of. If the material is not valuable, 
the film should be disposed of at once. Adhesiveness prevailing in de- 

Fig. 4. Sample of film in the third stage of decomposition. 

teriorating film may cause the emulsion to become detached from the 
base while unrolling. This frequently can be prevented by slowly un- 
rolling the film in a bath of carbon tetrachloride under precise labora- 
tory control. This should be done only in adequately ventilated 
areas. In the third stage, only small portions of the film may be re- 
producible. The reproducible portions should be separated, if valu- 
able, from the rest and copied. After reproduction, the entire original 
film should be immediately destroyed. In the fourth and fifth stages, 
film is photographically worthless and should be destroyed at once 
without further consideration. 




Fig. 5. Sample of film in the fourth stage of decomposition. 

Fig. 6. Sample of film in the fifth stage of decomposition. 



Films of stages three, four, and five, designated for disposal, 
should be immediately submerged in water-filled drums. They 
should be carried in these drums to a remote, uninhabited area ap- 
proved by fire authorities and destroyed by burning. The ground on 
which the film is to be burned should be free of brush, grass, leaves, 
and combustible litter. Burning should be confined to batches of not 
more than 25 Ib, as the heat from the burning of large amounts of film 
creates a strong updraft which may bear fragments of burning film 
considerable distances to endanger neighboring properties. Under no 
circumstances should films be burned in an inhabited area or within a 
building. The rapid production of gases by burning film makes it ex- 
tremely dangerous, particularly if burned in a furnace or confined 
space. During test fires in a well vented vault, engineers of the Inter- 
agency Advisory Committee for Nitrate Film Vault Tests have re- 
corded pressures as high as 18 psi. It is readily understandable that 
no ordinary furnace structure could withstand this pressure; its 
breeching would fail and fill the furnace room with flames and poison- 
ous gases. 


It is quite possible in the initial inspection that a relatively high 
proportion of film in advanced stages of decomposition may be found. 
The opening of cans containing this film may liberate quantities of 
noxious gases into the working area. Personnel exposed to them may 
experience nausea, headache, and other unpleasant symptoms if the 
ventilation is inadequate. It is, therefore, recommended that the per- 
sonnel working on old film inspection be given ten-minute rest periods 
each hour in the outer air. 

If we are to enjoy freedom from film fires during the coming sum- 
mer, a comprehensive program of film inspection should begin now 
so that the task may be completed before the onset of hot weather. 
Since film is constantly subject to decomposition, inspection should be 
repeated annually, preferably in the spring. Only by such procedure 
can we avoid the insidious menace to life and property hidden in de- 
teriorating motion picture film. Particular attention should be given 
to film of unknown quality or of obscure origin. 

Sensitometric Aspects of 
Background Process Photography 



Summary The composite negative obtained by photographing action 
against a rear projected background is a combination of an original negative 
with a dupe negative. An analysis by sensitometric methods of the grada- 
tional distortions thereby introduced has been attempted, the results of which 
are discussed in this paper. 

THE advantages of background process photography over straight 
production technique consist generally in savings in location 
costs and in permitting a convincing picturization of supernatural 
effects. The same is true of a number of other photographic effect 
methods, all classified under composite photography, of which differ- 
ent types of matte processes are examples. A singular point which 
background process photography can claim as one of its distinguishing 
features is that the composite effect of background and foreground 
can be observed and evaluated prior to and during the actual photo- 
graphing of the composite scene. As it presently stands, background 
process photography is but a small fraction of general motion picture 
production due to yet unsolved technical deficiencies and also on 
account of a somewhat reputed unpredictability of results. The 
combined advantages offered by this process are so obvious and sub- 
stantial that its increased application seems highly desirable. 

An investigation of the problems limiting the usefulness of back- 
ground process photography in motion picture production has become 
one of the major projects of the Motion Picture Research Council. 
This rather basic study has not yet substantially advanced past the 
required preliminary work of evaluating available literature and tech- 
nical knowledge represented by a store of practical experience in the 
studio transparency departments. This paper, therefore, is primarily 
confined to an analysis of some of the factors controlling the pictorial 
quality of the final composite obtainable with present methods. 
Such an analysis may contribute to the development of better meth- 

PRESENTED: October 14, 1949, at the SMPE Convention in Hollywood. 



ods and equipment which eventually should make use of all the poten- 
tial aspects expected of this process. 

In parallel to the historic development of other phases of photog- 
raphy, background process technique has advanced, so far, on a 
purely empirical basis to a remarkable degree of performance. To 
develop this technique to greater efficiency, dependability and stand- 
ard of quality, it becomes necessary to give it a clear and precise 
theoretical foundation. 

Our discussion here is confined to black-and-white technique. It 
is further limited to the consideration of gradational requirements and 
problems of background process photography. This, therefore, ex- 
cludes a treatment of such factors as definition, resolving power, 
graininess, etc., which are, of course, of vital importance in relation 
to the quality of the final pictorial reproduction. 


An analysis of the possible sources of gradational distortions reveals 
three major causes: 

1 . Duplicating Distortions 

This type of degradation is an inherent property of any duplicating 
process, bringing about a shortening of the straight line part of the 
characteristic curve. It may, therefore, affect gradational qualities 
of the background image reproduced in the final composite negative 
or print and will do so equally over its total area. Duplicating tech- 
nique in straight printing as performed in processing plants enjoys a 
relatively comfortable latitude in the variation of density levels and 
gamma values of the intermediate print and of the dupe negative to 
keep gradational distortions in the dupe positive to a minimum. 
Background process photography, however, is restricted to a much 
greater extent principally by the fact that the processing of a com- 
posite can give consideration only to an optimum gradation of one 
part of the composite. This means invariably favoring the gradation 
of the original foregound action negative. 

2. Distortions Affecting Different Background Image Areas 

This type of disturbance is evident in density and contrast varia- 
tions between the central image areas and off-center areas of the 
background component. They are caused by various factors, not 
easily isolated for the determination of each one's specific contribu- 


tion. Individually and in combination they reduce the image-form- 
ing illumination incident on the film progressively with increasing 

3. Distortions Caused by Flare 

Flare is non-image-forming illumination incident upon the film 
during its exposure in the camera. If present in sufficient intensity 
it will cause degrading of contrast and brightness range. The most 
common source of flare encountered in background process photog- 
raphy is "leak" light reflected from foreground objects onto the 
process screen. Other causes, however, are suspected of contributing 
substantially to this type of degradation. 


Ideal photographic reproduction of the projected background image 
requires by theoretical postulation linear density-brightness charac- 
teristics and an effective slope or gamma equal to unity. That this 
condition is not fulfilled in practice would, within reason, be just as 
acceptable as the common degradations in production release prints 
which probably never have straight line characteristics from clearest 
high light to deepest shadow. The adaptability of our aesthetic 
faculties apparently tolerates considerable deviations. However, as 
equally known from experience with other sensual observations, criti- 
cal judgment becomes immediately acute in case a basis for a direct 
comparison of two even slightly different reproductions is presented 
to the eye. This is the case with composite prints in which the grada- 
tion of the foreground portion varies from that of the background 


A brief outline of the principles of background process photography 
as practiced in studios is given here as an introduction to the later 
treatment of its gradational problems. 

The original background negatives are usually photographed on 
locations selected for this purpose, in which case the use of a panchro- 
matic, fine grain emulsion type is generally adopted. Sometimes, 
stock shots of older date and not originally intended to serve as 
background negatives have to be used. In either case, a print is 
made, normally in contact, of such negatives which becomes known 


as the "process plate." The positive stock used for making this 
print is in most instances regular release print emulsion type with a 
normal positive gamma. A special fine grain print stock with con- 
siderably lower gamma characteristics is also used. The process 
plate is projected in a special background projector onto a translucent 
screen. The rear projected image is photographed together with 
the action in the foreground by a standard black-and-white camera. 


Fig. 1. Optical test arrangement (schematic). 

Figure 1 is a schematic illustration of the optical test arrangement 
used in our experiments in investigating small process screen samples 
of different makes. An explanatory guide to the letter symbols used 

a Projector film gate 

ab Focal length of projector lens = 5 in. 

c 8 X 10 in. screen 

ck Horizontal screen dimension = 10 in. 

bm Distance between screen c center and projector lens = 3 ft lOVs in. 

mg Distance between screen c center and camera lens = 1 f t 8 36 / 6 4 in. 

gh Focal length of camera lens = 2 in. 

h Camera film gate 

d 30 X 40 in. screen 

dl Horizontal screen dimension = 40 in. 

m'b Distance between screen d center and projector lens = 15 ft 4*/2 in. 

m'e Distance between screen d center and camera lens = 6 ft 10 in. 

/Focal length of camera lens = 2 in. 
Camera film gate 

This arrangement was designed in exact correlation to the system 
commonly used in studio technique. Naturally, the figures for the 
distance between screen and projector and equally between screen 
and camera will increase proportionately with the size of the process 
screen employed. The projector and camera half-angles used in this 
illustration are co-ordinated to the focal length and aperture of the 
projector lens and of the camera lens respectively: 

Projector half-angle, vertical: 4 Camera half-angle, vertical: 9 

Projector half-angle, horizontal: 5.5 Camera half-angle, horizontal: 12V2 
Projector half-angle, diagonal: 7 Camera half-angle, diagonal: IS 1 /* 

Use of a 5-in. projection lens and a 2-in. camera lens is representative 
of the most severe conditions under which background process pho- 




tography is normally practiced. Wherever possible, lenses of greater 
focal length are employed to keep projection and camera angles as 
small as possible. 

The gradational distortions to be discussed may be inherent in the 
"dupe negative"* alone or also in the process plate, both being the 
photographic media involved. The transmission characteristics of 
the process screen are all-important as factors influencing the grada- 
tional properties of the dupe negative. It therefore seems practical 
to treat briefly the photographic characteristics of the process plate 
separately and also to give special attention to the physical character- 
istics of diffusely transmitting objects in the form of the presently 
used translucent background screens. 


Figure 2 shows the gradational characteristics of a sensitometric 
negative strip and of a contact print made from this negative. Ex- 

Fig. 2. 

Intensity scale (factor \/2) 
Original Negative (O.N.) 
Contact Print (C.P.) 
Negative density range (str. line), 

Print density range (str. line), 

Negative exposure latitude (str. 

line), 1:56 
Print exposure latitude (str. line), 


(LOG) E 

posure and processing for both original negative and contact print 
were closely matched to conditions which govern the making of 
process plates. 

It will be noted that the linear portion of the printed curve is rela- 
tively short, being terminated at the lower densities by the character- 
istic curvature of the positive material, and at the higher densities 
by the inherent curvature of the negative material. 

The over-all gamma value of 1.31 for the print is considerably above 
the postulated value of unity. This may be necessary for two reasons, 

* By "dupe negative," as used from here on, is meant the rephotographed 
negative of the projected background plate. "Dupe print" signifies contact 
print from "dupe negative." 




one of which is to obtain an over-all contrast close to unity, the other 
to compensate for contrast reducing factors encountered in projecting 
and rephotographing the process plate from the process screen. This 
would mean that to obtain an effective contrast value of unity, the 
process plate has to be developed to an over-all gamma of 1.31. 

From the graphs it may further be concluded that an increase in 
print exposure which would shift the position of the print curve to 
the right should result in an expansion of its linear part, while a de- 
crease in exposure will cause it to compress further. This would indi- 
cate that heavy prints are preferable for process plates. Insufficient 
projection illumination, however, limits the use of high density prints 
in many practical setups. 

The background process screens used at present are translucent 

media to be classified as imperfect diffusers. 

The brightness or luminous intensity distribution over the area of 

such a screen type differs widely from that of a perfect diffuser that 

follows the simple cosine law of emission : 

/ (0) = 7o X Cosine 0, 

where, IQ = Candlepower of the surface normal to itself (6 = 0) 
I (8) = Candlepower at an angle 6 from the normal. 

An illustration 1 of the difference in transmission characteristics for 
various diffusing materials in comparison to the cosine function of a 
perfect diffuser is given in Fig. 3. 

10 20 30 40 50 60 70 80 90 

Fig. 3. Transmission characteristics 
of glasses. 

0, Cosine curve 

1, Solid opal glass depolished 

2, Flashed opal glass, acid etched 

3, Flashed opal glass, natural 

4, Flashed opal glass, sandblasted 

5, Clear glass, both faces sandblasted 

6, Clear glass, one face sandblasted 

7, Clear glass, one face acid etched 

The rapid fall-off with increasing angle typical of imperfect dif- 
fusers is the main cause of the poor brightness distribution of pres- 
ently available process screens. Unfortunately, the closer the trans- 
lucent medium approaches the characteristics of a perfect diffuser, 




the lower, in general, will be its relative transmission. In other 
words, gain in wide angle brightness distribution is accompanied 
by a loss in relative transmission or an increase in opacity. Figure 4 
illustrates this dependence between transmission / and 1(6) at a 
fixed angle of = 10 for 
several brands of commer- 
cial glass luminants with 
widely varying diffusing 

The angles encountered 
under practical conditions 
in 35-mm background 
process photography are 
within 15 for the projec- 
tor side and 30 for the 
camera. The extreme 
angles stated are, however, 
quite frequently used of 
necessity. In some cases 


-P 90 

g 30 
tf 70 
-f 60 




D 40 



15 30 45 60 75 90 

Fig. 4. 

1, Factrolite 

2, Tapestry 

3, Maze 

4, Sandblasted 

5, Flashed opal 

these angles may be still 
enlarged by the use of 
shorter focal length objec- 
tives, particularly for projection, when space limitations may leave 
no other choice. 

The process screen type in use at present consists of a transparent 
cellulose acetate sheet which is sprayed on one or both surfaces with 
a layer of zinc stearate suspended in a solution of cellulose acetate. 
These layers serve as the diffusing medium which is, of course, needed 
to stop the projected image on the screen and make it visible and 
reproducible by photography. Each screen is handmade, and uni- 
formity of transmission characteristics over the total area and from 
one screen to the other is therefore practically impossible to attain. 
The surfacing is accomplished by hand spray techniques which cannot 
assure even thickness of the coatings. The problem of uniformity 
becomes obviously more serious the larger the screen size. The trans- 
mission factor for = is determined in studio practice by simple 
comparative photometric readings of the amount of incident and 
transmitted light and serves as an indicator of the "speed" of the 
screen. Transmission readings made on a considerable number of 
screens at zero angle indicate that the luminous intensity of studio 


screens varies in the range of from 55 to 70 per cent of the intensity 
incident upon the screen. The transmission characteristics deter- 
mined by light meter readings are usually checked by means of a 
photographic density test. Such transmission readings serve in 
practice for guidance in arriving at a proper balance between level 
of screen brightness and foreground illumination. They are apt, 
however, to lead to erroneous calculations since they do not differen- 
tiate between specular and diffused transmission, and may also in- 
clude light rays which do not reach the camera lens. 

The present limitations in maximum light output of the background 
projector light source and optics make it necessary to increase the 
transmission of screens in direct proportion to their size, which the 
screen maker attempts by varying his spraying technique. Since it 
is desirable for various reasons to keep the light output of the pro- 
jector constant, it is common practice to use denser process plates 
for smaller screen sizes. This means that for small screens with a 
relatively short throw, process plates of comparatively high density 
and low gamma are used, while for large screens and long throws, 
prints of relatively low density and high gamma are preferred. 
The over-all contrast of such different print types does not, of course, 
necessarily change by this procedure. 

The transmission and light scattering characteristics of translucent 
screens have a large and multiple influence on the gradational char- 
acteristics of the composite photograph that is the end product of the 
background projection process. As it stands at present, practically 
nothing has been done toward a quantitative evaluation of these 
factors, with methods taking into specific consideration this field of 
background process photography. Much of the data available 
from literature on the subject of scattering of light 2 can be directly 
or indirectly utilized to great advantage in placing background process 
photography on a more efficient and reliable basis. 

In addition, however, an extensive experimental investigation of 
materials and methods potentially suitable for making improved 
screens is indicated. Since the complexity of transmission character- 
istics of imperfectly diffusing media prevents the application of a 
clear mathematical treatment and since an additional photographic 
phase contributes to these difficulties, it seems fairly certain that 
direct practical experimentation studying size, refracting index, 
opacity, shape, distribution and other factors on a great number of 
materials will become unavoidable. 


Practical experience in background process photography has, of 
course, established the validity of certain fundamentals which serve 
at present as rather general guides for screen makers and screen users. 
One of these, for instance, is the well-known observation that increase 
in diffusion decreases definition and contrast. 

1 . Sensitometric Investigation of Typical Studio Procedure 

The original sensitometric negative strip and print referred to earlier 
and shown in Fig. 2 were produced for this experiment in the follow- 
ing manner. A studio 35-mm motion picture camera with single 
frame exposure device was aligned on an optical bench with a point 
light source as a photographic target. Single frame exposures were 
made in succession with full aperture and at a constant speed of 
Y% sec. The distance of the light source from the camera was in- 
creased by precalculated values to result in exposures with inten- 
sities decreasing by factor ^2. The negative film type used was 
Eastman background emulsion. Processing by machine development 
was performed by Pathe* Film Laboratories with instructions to de- 
velop to a customary negative gamma of 0.71. The developed nega- 
tive was printed in contact onto Eastman Positive 1302 film and the 
resulting print developed for a control gamma of 2.15. 

This print of an intensity scale negative strip was projected and 
rephotographed in regular background process procedure on a stage 
of the Universal-International Studios, together with pictorial process 
plates, and therefore is representative of the technique as presently 
practiced. The only exception to actual production conditions in 
background process photography consisted of omitting the inclusion 
of flare light from front illumination. The print was projected by a 
Mitchell background projector with a lens of 5-in. focal length onto a 
9 X 12 ft process screen of the Sanders type from a distance permitting 
the projected image to fill the screen area. 

The screen image was rephotographed by a standard 35-mm camera 
with a lens of 2-in. focal length and an aperture ratio of //4. The 
camera was aligned with the projector and the center of the screen. 
The maximum off-axis angle formed by the projector was 7, that 
formed by the camera was 15. The resulting dupe negative was 
processed by machine development to a normal negative action 
control gamma of 0.67. A contact print from this negative was made 




onto Eastman Positive Type 1302 on Light 11 and processed to a 
positive control gamma of 2.15. Figures 5 and 6 respectively show 
the density-log exposure graphs corresponding to both dupe negative 
and dupe print. Each of these figures depicts curves numbered from 
1 to 4 which were obtained from density readings corresponding to 
various angles of illumination incident upon the film during exposure 
in the camera. It is quite apparent that for extreme off-axis image 
details, not merely a reduction in negative densities and a correspond- 
ing increase in print densities takes place, but also a pronounced 
change in curve shape and gamma compared with a curve constructed 
from the density readings of the image center. 

(LOG) E 

(LOG) E 

Fig. 5. Studio dupe negative. 

1, Center readings 

2, 5 angle off center readings 

Fig. 6. Studio dupe print. 

3, 10 angle off center readings 

4, 15 angle off center readings 

The progressive decrease in image-forming illumination on the film 
with increasing angles of incidence, which in turn resulted in the 
gradational differences shown in the above example, is mainly caused 
by the brightness fall-off characteristics of the process screen used. 

The making of the process plate could not involve major off-axis 
distortions since in exposing the original negative the object was 
located on the optical axis so that vignetting and barrel effects as well 
as cosine 4 factor did not vary the image-forming illumination to any 
appreciable extent. 

The background projector optics in this instance are adjusted to 
give a practically flat field for a half angle of 7, and light meter read- 
ings of the radiation incident upon the screen did not vary more than 
5 per cent from center to corner. 

In rephotographing the projected image, stop//4 was used which at 
least eliminates vignetting and barrel effects. 


2. Sensilometric Testing of Different Types and Makes of Process 

A number of screen samples secured from three local fabricators 
were tested sensitometrically to determine and compare the influence 
of their transmission characteristics upon the gradational properties 
of photographic reproductions made in background process procedure. 
The test arrangement was similar to the one described and illustrated 
in Fig. 1. For the sake of simplicity and economy, and to obtain re- 
producible results in check tests, the following deviations from studio 
procedure seemed permissible. 

As a projector, a factory-pretested Kodaslide still projector was used 
with an//2.3 coated Ektar lens of 5-in. focal length. The illumination 
consisted of a 500-watt incandescent lamp operated at a voltage of 
115 volts d-c which was carefully maintained throughout the tests. 

The screen samples were uniformly kept to an 8 X 10 in. image area. 

To obtain the process plate for this test, an 8 X 10 in. paper print 
of a step-wedge with six progressively increasing densities was photo- 
graphed with a Contax camera on Eastman Background X type film 
and this negative developed to a gamma of approximate unity. 
The exposure was adjusted in such a way that the darkest field of 
the original step chart reproduced on the film with an approximate 
density of 0.40, the lightest field with an approximate density of 1.6 
so that part of the film record registered on the lower curved portion 
of the characteristic curve of the film material. The image dimen- 
sions of the step-wedge reproduced on this film covered an area of 
only .25 X .06 in. in the center of the Contax picture frame. This 
frame was masked down to one-half standard 35-mm background pro- 
jector aperture and projected in the Kodaslide projector onto the 
different 8 X 10 in. screen samples. At a distance of 69.5 in. between 
screen and projector, the screen image of the wedge extended length- 
wise to the horizontal edges of the 8 X 10 in. screen. All stray light 
was masked off. This arrangement permitted projection with a maxi- 
mum of central rays. The screen image was rephotographed from 
the back with an 8 X 10 in. view camera using an Ektar lens with 
12-in. focal length and a 4 X 5 in. back. The film type selected for 
this purpose was Panatomic Cut Film. The film records were devel- 
oped to a control gamma of 0.70. Exposure was kept constant for all 
tests and adjusted to the medium brightness level of all screens. 

In rephotographing the projected screen images the camera was 
placed in two positions as follows: 


A. Camera aligned with screen center and optical axis of projector. 
Distance between camera and screen center, 20 in. Placement of 
screen surface normal to optical axis of projector. 

B. Camera moved parallel to screen with lens focused on center 
of screen until angle of 12^ between optical axis of camera lens and 
normal to screen was formed. Distance between camera and screen 
center, 21.5 in. Lens refocused for this distance. Placement of 
screen surface normal to optical axis of projector. 

While this arrangement contains a number of unknown factors and 
does not permit a quantitative evaluation of the results, it was felt 
adequate in providing reliable comparative information which was 
the primary aim of this test. 

V v-- 

I7.5VX^3 L- 7 c 

'>< ' 

/ 12.5 \ c Fig. 7. (Not to scale.) 

Figure 7 is a schematic illustration (angles not drawn to scale) of 
these two setups wherein C depicts camera position A, and Ci camera 
position B. In both instances the optical axis of the camera lens 
meets the optical axis of the projector lens at the center of the 
screen S. 

The B-shot permits evaluation of transmission fall-off for a 12^ 
angle, characteristic of each screen. The arrangement was chosen 
to minimize factors causing additional brightness fall-off due to the 
projector optics employed. Naturally a slide projector of the type 
used does not give a similarly flat field as a specially corrected studio 

The step densities of each film record were read and the readings 
applied to construct density exposure curves. Since the factor of 
exposure progression was not known, the exposure scale is entirely 




Table I lists for each screen and each camera setup the gamma read- 
ings and the density values of Step 1 (lowest density), Step 4 (me- 
dium density), and Step 6 (highest density). The gamma readings 
have, of course, no quantitative values. They permit, however, a 
comparative evaluation of the influence of each screen on contrast. 


Screen Type Readings 

1st Step 

4th Step 

6th Step 
































































































































Double 1.49 










Double = double surfaced; Single = single surfaced. 


(LOG) E 


(IOC) E 



(LOG) E 


2 3 4 4 

(LOG) E 


(LOG) E 

Fig. 8. 

(LOG) E 


Figure 8 shows graphically these density curves for six selected 
screen types corresponding to Screens No. 1 to 6 of Table I. The 
solid line curves are constructed from densities obtained with A-shots 
(camera on optical center). The dotted line curves were plotted from 
densities obtained with B-shots (camera 12J^ angular position). As 
will be noticed, the six selected screen types are progressively worse 
in fall-off characteristics. 

In studying and analyzing the data contained in Table I, a clear 
relation between fall-off characteristics and relative transmission of 
the screens could not be established, nor was it possible to find a 
simple function of contrast in relation to relative transmission. This 
result is rather disappointing and emphasizes the complexity of factors 
under which background process photography has to be practiced at 
present. The irregularities observed, however, make it practically 
certain, in an indirect way, that more efficient screens can be con- 
structed, that is, screens with optimum relative transmission (speed) 
and still relatively favorable angular fall-off characteristics. 


The tests described in this paper were primarily selected out of a 
rather large series of similar tests performed in the course of this proj- 
ect, to exemplify the type of studies that have so far been undertaken. 
They further illustrate the problems that have been encountered and 
are expected to be faced in continuation of this work. Finally, they 
indicate the manifold factors and phases that are left for further 
experimentation and analysis. 

While improvements of present type process screens and process 
technique appear entirely feasible, they may not lead to ultimately 
satisfactory results unless the complications and limitations of imper- 
fectly diffusing screens are eliminated. This indicates a search for 
different types of projection screens which, while not included in the 
present discussion, is also under continuous consideration by the 
Research Council. 

At this stage it is too early to make any suggestions or predictions 
concerning the development of basically new process screens and 
techniques. An investigation into the potential application of 
directional screens has been undertaken as part of our general proj- 
ect. However, the large size screens demanded for motion picture 
production work and the trend of making increased use of camera 
movement in front of the projected background present severe diffi- 


culties and limitations in the manufacture and use of directional 


We wish to express our appreciation to Universal-International 
Pictures Inc. and 20th Century-Fox Film Corp. for their liberal 
assistance extended to the Research Council in the staging and process- 
ing of our experimental work. We also acknowledge valuable infor- 
mation and suggestions given to us by Farciot Edouart and Dr. 
Charles Daily of Paramount Pictures Inc. and William Slaughter 
of Metro-Goldwyn-Mayer Studios. Special mention is due Stanley 
Horsley, of Universal-International Pictures, for his untiring efforts 
and interest in this project. He also guided the design of our test 
equipment and supervised the control tests made at that studio. 


(1) M. Cohu, C.I.E. Proc., pp. 440, 445; 1931. 

(2) Literature on diffusion, scattering and miscellaneous subjects discussed in 
this paper: 

Lord Raleigh, Proc., Royal Soc. of London, vol. 90, p. 219; 1914. 

C. V. Raman and B. Ray, Proc., Royal Soc. of London, vol. 100, pp. 102-9; 

J. W. Ryde and B. S. Cooper, Proc., Royal Soc. of London, vol. 131, p. 451; 

"Report of I.E.S. Committee on Illuminating Glasses," 1933. 

"I.E.S. Report of Subcommittee on Diffusion," p. 109; 1939. 

A. Paulus and C. S. Woodside, I.E.S. Trans., vol. 9, p. 749; 1933. 

M. H. Bigelow and A. M. Howald, I.E.S. Trans., vol. 4, p. 414; 1936. 

F. W. Warner, I.E.S. Trans., vol. 3, p. 244; 1938. 

M. H. Bigelow and A. F. Wakefield, I.E.S. Trans., vol. 10, p. 1189; 1939. 

"Bibliography of Papers on Diffusion," /. Opt. Soc. Amer., vol. 1, p. 1230; 

J. W. Ryde and D. E. Yates, /. Soc. Glass Tech., vol. 10, p. 274; 1926. 

Arthur C. Hardy, /. Frank. Inst., vol. 204, p. 783; 1929. 

J. W. Ryde and B. S. Cooper, Proc., Int. Com. Ilium.. 1931. 

R. S. Stein and P. Dotz, J. Amer. Chem. Soc., vol. 68, p. 159; 1946. 

P. P. Debye, J. Phys. & Coll. Chem., vol. 51, p. 18; 1947. 

P. P. Debye, Chem. Rev., vol. 43, p. 319; 1948. 

A Motion Repeating System 
For Special Effect Photography 

BY 0. L. DUPY 


Summary This is a motion reproducing system for use in "special effect 
shots' 1 (multiple exposures, matte paintings, pan and tilt traveling matte 
shots, etc.) which permits the change of a camera position and adjustment 
while photographing a scene. Then, with the film rewound to the original 
starting point, a second exposure is made on the same film of a second scene, 
duplicating the relation of camera position and adjustments to the film 
movement which existed while photographing the first scene. This is done 
with sufficient accuracy to match matte lines in any of the standard double- 
exposure or other similar effect shots. 

T~T TITH MODERN STUDIO EQUIPMENT the camera is free to move about, 
VV follow the action, and graphically and pleasingly describe the 
scenery and tell a story. But all this freedom of camera movement 
does not help the makers of special photographic effects, because some 
processes require that the camera be fixed, as the film is run through 
the camera several times and the image of one exposure must be 
positioned in exact relation to the other parts of the picture pre- 
viously exposed. When cut into a picture that has been photographed 
by a roving camera, the effect shot, by comparison, may seem static 
and lifeless regardless of how well conceived and executed. 

Some double exposure and matte shots have been made with the 
camera moving mechanism and the camera film-moving mechanism 
geared together. By beginning each take with the film and camera 
positioned on a start mark, the camera position will be the same for 
each picture frame, as the film passes through the camera for any 
number of operations. The use of this system is limited because the 
rate of movement must be predetermined and the action must follow 
the camera, which is an awkward, and sometimes an impossible, pro- 

A system for special effect shots has been devised and applied at 
present to panning and tilting the camera, which permits the camera- 
man to pan and tilt the camera in a normal manner and follow 
the action as desired. A record is made of the movement and, for 
subsequent exposures on the same film, the record controls the camera 
PRESENTED: October 14, 1949, at the SMPE Convention in Hollywood. 


movement, matching the original relation between the camera posi- 
tion and picture frame during these subsequent exposures. 

In this system the camera is moved by the output shaft of a dif- 
ferential gear system. Two synchronous motors are connected to 
this gear system in such a way that the output shaft rotates at a 
speed which is the difference in the speed of the two motors; in other 
words, if the two motors were connected to the same power source, 
the difference in speed would be zero and consequently the speed of 
the output shaft, which moves the camera, is zero and the camera 
is stationary. However, if one of the motors is connected to a power 
supply of a different frequency, the camera is moved at a rate that 
is in proportion to the difference of the frequency of the two sources. 

In operation, one of the synchronous motors is connected to the 
electric power system and runs in synchronism with the power line 
frequency; the other, whose change in speed causes the camera to 
move, is fed by an amplifier and runs in synchronism with a control 
signal which is fed to this amplifier. This control signal is produced 
by an induction frequency changer. The three-phase primary of 
this frequency changer is also connected to the electric power system. 
With this type of frequency changer, the output frequency is the same 
as the input when the rotor is at standstill and increases in proportion 
to the speed of rotation if the rotor is revolved in one direction and 
decreases proportionately if reversed. Therefore, the camera move- 
ment follows the rotation of the frequency changer shaft. 

This control signal is also fed to a recording system and recorded 
when the first take is made. 

A "start mark" is placed at the beginning of the record and the 
beginning of the picture film. The recording machine and the camera 
are driven by a common-drive source so that they accelerate and run 
together and the length of the picture film and the record are in pro- 
portion through the running. 

For subsequent running of the film through the camera, the camera 
and record-reproducing machine are lined up with their respective 
start^narks as they were at the beginning of the first running and are 
driven by a common-drive system. The signal, which is fed to the 
amplifier driving the variable-speed synchronous motor, is reproduced 
from the record. Being identical in respect to frequency to the orig- 
inal signal from the frequency changer that produced the camera 
movement during the first run, the camera moves as it did during the 
first run. Thus, as each picture frame is moved into the photo- 




graphic aperture, the camera is in the same position and angle 
as it was when that picture frame passed through the camera the 
first time. The camera is held in a fixed position for a short time at 
the beginning of the shooting. On subsequent running of the film 
this time interval is used to check and correct the camera position. 




Fig. 2. Motion repeating machine for Technicolor camera, 
with Bell & Howell camera adaption. 

-4 ~t 


Fig. 3. Recording and reproducing equipment and controls, 
with all electronic equipment mounted in portable cabinet. 

294 0. L. DUPY 

The phase angle of the signal from the disc may vary a few degrees 
from the original signal; so after everything is up to speed this error 
is corrected by means of a phase shifter which is inserted in the con- 
stant-speed synchronous-motor line. 

The common-drive system that is mentioned is the same electrical 
interlock system that is used in sound picture work to hold the picture 
and sound systems in synchronism for sound picture making and re- 
production. It is also used to hold the camera and projection ma- 
chine shutters and film movements in synchronism for making pro- 
jected background process shots. 

Figure 1 is a block schematic of the system showing, for simplicity, 
one movement only for panning the camera. A duplicate system is 
provided to elevate and depress the camera. Both records are made 
simultaneously on one disc. A system similar to the standard 
phonograph disc recording and reproducing system was used to record 
and reproduce the control signal. 

The power amplifier which supplies the power to the variable-speed 
synchronous motor is in this application a thyratron inverter. This 
system is ideal for this job because variations in input voltage have 
no effect on the output voltage, provided the input voltage stays 
above the minimum necessary to fire the thyratrons. Also, the simple 
circuit used produces the necessary change of output voltage with 
frequency to operate a synchronous motor, at full torque, at all 
frequencies. Ample power is produced using the 120 d-c studio 
lighting power as a "B" supply. 

Errors due to the shifting of synchronous-motor rotor poles, in 
respect to the applied frequency as it takes a load, are reduced by a 
factor of 800 to 1 in the gear train. A large portion of this displace- 
ment occurs when the motor accelerates or decelerates the camera, 
and as the motor has the same mass to move at the same point of its 
movement no visible error in duplication of movement due to this 
has been observed at any time. The present maximum rate of pan 
and tilt speeds is about 25 degrees per second, which is faster than 
any requirement encountered at present. 

Mr. Richard Duval of the Development Engineering Department 
of Metro-Goldwyn-Mayer Studios has worked out lens position 
charts for the changing of focal settings between various shots. 

Using a painting as the second subject photographed, a change 
of ten to one in focal distance has been made with no relative move- 
ment between the images photographed at these two settings. 

Increased Noise Reduction 
By Delay Networks 


Summary This paper describes a new method of obtaining increased 
signal-to-noise ratio in optical sound film recording. This is done by in- 
creasing the noise reduction and is made possible by the use of delay net- 
works which delay the application of sound currents to the modulator until 
after the noise-re duction-bias current has been partially canceled. Noise- 
reduction settings as high as 30 db have been tried with good success and 
settings of 15 db have been used in regular production. 

RECENT DEVELOPMENTS in original sound recording such as 200- 
mil push-pull optical tracks and magnetic films have made 
original sound records far superior to the release tracks which are 
heard in the theater. Since it undoubtedly will be many years before 
anything except the standard non-push-pull optical sound tracks will 
be used in theaters, anything which will improve the quality of release 
tracks would be highly desirable. 

One relatively easy way of improving the release track would be 
to increase the noise reduction now being used in all optical film re- 
cording. 1 However the use of noise reduction brings in some inherent 
disadvantages such as "clipping" and "thumping." Clipping is 
caused by the modulator overloading for a few syllables when sounds 
with sudden impacts are being recorded, because the cancellation of the 
noise-reduction-bias current takes an appreciable length of time. If 
this time is shortened too much in an effort to reduce clipping then the 
operation of the modulator by the noise-reduction currents approaches 
audible frequencies and thumping occurs. In practice the noise 
reduction unit is adjusted for a practical balance between clipping 
and thumping which together with the use of "margin" makes these 
sounds generally unnoticeable. Margin is the adjustment which 
determines the difference in decibels between the signal which just 
overloads the modulator and the signal which just cancels the bias 

One method of reducing these undesirable noise-reduction effects 

PRESENTED: October 14, 1949, at the SMPE Convention in Hollywood. 





so that increased noise reduction could be used would be to delay 
the application of the signal currents to the modulator until after the 
noise-reduction currents were partially canceled. Up to the present 
time this has not been practicable but with the newly developed 
Delay Network designed by the Bell Telephone Laboratories it has 
now become possible. 


Fig. 1. Delay network rack, front and rear views. 

For several years delay networks have been used in radio receiving 
systems, automatic recording oscillographs, transoceanic radio tele- 
phone service and radar. 2 The design of delay networks is based 
on the principle that electrical energy supplied to inductances and 
condensers can be stored for an appreciable length of time in the 
electromagnetic and electrostatic fields of the coils and condensers. 
This storage creates a time delay between the receipt of a signal at 




the input to the network and its delivery to the output. If the net- 
work is designed to have different amounts of delay at different fre- 
quencies it is called a delay equalizer; while if it is designed to have a 
constant delay for all frequencies it is usually referred to as a delay 
network. It is relatively easy to design a delay network to cover a 
limited band of frequencies but it is very difficult to design one which 
will cover the full audio-frequency band. The new networks are 
designed to cover the frequency range of 50 to 8000 cycles and have 
been made possible only by the great amount of experience gained 
by the Bell Telephone Laboratories in recent years in designing delay 
networks. Each network has a constant attenuation of about 
13.5 db and a time delay of 8^2 milliseconds. Delay times in mul- 
tiples of 3^2 milliseconds can be obtained by using the networks in 

Fig. 2. Delay network rack, block diagram. 

For reasons explained later it was decided that in production sound 
recording, four networks in series, having a time delay of 14 milli- 
seconds, would be used. These networks, together with the neces- 
sary amplifiers, power units and attenuators, were assembled on a 
rack (Figs. 1 and 2). The delay network rack circuit was designed 
to have a zero insertion loss and enough output capacity to drive a 
light-valve modulator. On the recommendation of the Bell Tele- 
phone Laboratories the circuit was so designed that no network would 
have an input power exceeding zero dbm. Following good engi- 
neering practice, the networks are isolated by attenuators from possi- 
ble variations in impedances. As shown in Fig. 3 the delay network 
rack is patched into the re-recording channel just ahead of the modu- 
lator and after the noise-reduction-amplifier input bridging point. 
In this way the speech currents are applied to the noise reduction 
amplifier 14 milliseconds before they arrive at the modulator. By 
























100 1000 



Fig. 4. Delay network rack, frequency characteristic. 



+ 5 -MO 


Fig. 5. Re-recording channel, harmonic distortion. 


having the network rack have a zero insertion loss, it can be removed 
or inserted in the channel without the necessity of changing the 
channel operating adjustments. 

The gain frequency characteristics of the network rack is within 
1.5 db over the range of 50 to 8000 cycles (Fig. 4). 

Distortion measurements made on the re-recording channel, which 
was a regular Western Electric 435 D re-recording channel using an 
RA-1231 B recorder and RA-1251 B re-recorders, show that including 
the delay networks the distortion for any frequency above 50 cycles 
is less than 1% for inputs up to 9 db above the light-valve overload 
point. At light- valve overload the distortion at the same frequencies 
is less than M% (Fig. 5). 

After considering the possible uses for delay networks in a sound 
recording channel, it was agreed that the most advantageous use 
would be to increase the signal-to-noise ratio in release prints. While 
the networks could be used to reduce undesirable noise-reduction 
effects such as clipping and thumping, it was believed that an in- 
creased release volume range was preferable. Consequently, our 
tests have been limited to that field. 

A normal noise-reduction setting for single variable-density re- 
cordings is 10 db with the noise-reduction filtering adjusted for an 
attack time of from 16 to 22 milliseconds. Attack time has been 
defined as the time required for the bias current to undergo 90% of 
its total change when a signal having a magnitude somewhat less 
than the value required to cancel the bias fully is applied to the noise- 
reduction unit. Since the movement of the light-valve ribbons is 
proportional to the amount of current flowing through them, the 
average spacing of the light valve will be directly related to the bias 
current. The characteristics of the noise-reduction timing filter on 
attack signals are such that the rate of opening of the biased light 
valve is not constant but is greatest at the beginning and decreases 
to zero at the end. For a given noise-reduction timing, the rate of 
opening varies with the noise-reduction setting, being faster for in- 
creased noise reductions. 

Since it was planned to use high values of noise reduction and since 
the rate of valve opening was to be retained at about the same value 
as was normally used, the timing filter was modified to have an attack 
time of about 24 milliseconds or about 25% longer than normal. With 
this filter and a noise-reduction setting of 15 db, the rate of valve 
opening is about the same as with a normal 19-millisecond filter and a 




setting of 10 db. With the modified filter, the valve opens to 41% of 
its average spacing in 14 milliseconds (Fig. 6). 

With a delay time of 14 milliseconds and with the modified filter, 
it was calculated that the valve could be completely closed and that 
less clipping of initial sound would occur than with normal operation 
at 10-db noise reduction. With this in mind, re-recording tests were 
made with music and dialogue with noise-reduction settings of from 
to 30 db and with margin settings from to 6 db. These tests 
showed that the use of high values of noise reduction was entirely 
feasible although as much as 30 db was undesirable since at that 
value the sound quality began to deteriorate somewhat. These tests 
also indicated that margin settings greater than zero were unnecessary 




Fig. 6. Sound tracks with and without delay networks. 

to prevent clipping, although the use of up to 6 db of margin caused 
no harm. However, the decrease in film noise made 96-cycle sprocket- 
hole modulation more apparent. For this reason it was decided to 
use a value of 15-db noise reduction for actual production re-recording. 

Sound Services, Inc., have been using this value of noise reduction 
for over a year with various types of product with completely satis- 
factory results. In a number of cases the increased noise reduction 
has been very helpful in allowing a decrease in the normal recording 
level so that special effects could be obtained. 

While most of our experience has been with density recording, we 
have also used the networks with success in variable-area recording. 
Here the problem is one of obtaining as narrow a bias line as possible 


without getting excessive clipping action. In order to reduce the 
low-frequency thump action which would have been accentuated by 
the increased rate of change of the noise-reduction ribbon due to 
the smaller valve spacing, it was necessary to increase the attack 
time of the noise-reduction filter as was done for density recording. 
An attack timing of 32 milliseconds was used. 3 With no other 
change, we were able to bias down to a bias line on the film of 1 to 
\\2 mils with good results. Experience with this small bias line 
has shown that films that have been projected many times in theaters 
still exhibit very little film noise. 

While a good deal of practical experience has been obtained in the 
use of 14 milliseconds of delay, no extensive investigations have been 
made regarding optimum time of delay, the best speed and type of 
noise-reduction attack timing, or the practical limits to the amount 
of noise reduction which can be used. Tests such as these should be 
made in the laboratory and the conclusions tried out under practical 
operating conditions. 


Delay networks with a time delay of 14 milliseconds over the audio- 
frequency range of from 50 to 8000 cycles have been used in sound 
film recording to obtain additional noise reduction. Through the use 
of these networks noise reductions up to 30 db have been tried and 
15 db has been used in production for a period of over a year. One 
of the limiting factors in using more noise reduction is sprocket-hole 
modulation which becomes more objectionable as noise reduction is 
increased. The use of the delay networks is advantageous in both 
variable area and variable density recording. They are especially 
desirable since the improvements gained through their use are obtained 
in the final release negative. 

NOTE: The presentation of this paper was supplemented by a demonstration 
film which was designed to show the possible uses of high noise reduction. 


(1) R. R. Scoville and W. L. Bell, "Design and use of noise reduction bias 
system," Jour. SMPE, vol. 38, pp. 125-147; February, 1942. 

(2) H. M. Thomsen, "New voice frequency electrical delay network," Bell 
LabsRec.,pp. 15-18; September, 1940. 

(3) Kurt Singer, "Versatile noise-reduction amplifier," Jour. SMPE, vol. 
50, p. 562; June, 1948. 

Miniature Condenser Microphone 



Summary This paper describes a miniature microphone that has been 
reduced in size so that an extremely uniform frequency response is obtained 
from all angles of pickup. Although it is very small its output level is high. 
Because it is so small, it can be used in many applications not possible with 
larger and heavier microphones. Uses of the microphone in motion picture 
sound recording indicate a different placement and pickup technique can be 
used. Having simple parts, its uniformity of production can be extremely 
high. A description of different models is given along with suggested 
methods of using the microphone. 

DOLBEAR described a condenser microphone in 1882 and such a 
device was reported as being shown at "La Lumiere Electrique" 
in 188 1. 1 Later Wente 2 produced a commercial unit which had ex- 
tensive application in the first years of sound motion picture record- 
ing. It was then replaced by dynamic, ribbon and directional micro- 
phones. Later smaller condenser microphones were described by 
H. C. Harrison and P. B. Flanders 3 and also by F. L. Hopper. 4 The 
principal objections to the early condenser microphone designs re- 
sulted from such factors as: (1) diffraction effect due to its compara- 
tively large size; (2) cavity resonance; and (3) electrical circuit com- 

In using a microphone the sound reaching it directly is, in most 
cases, only a very small part of the total. Most of the sound arrives 
from random directions, the amount being picked up directly depend- 
ing upon the reverberation of the room and the distance from the 
source. When the microphone has a response which differs in the 
various directions, the output fails to be an exact reproduction of the 
source. Under these conditions, one form of distortion is introduced 
by the microphone. The directivity of pressure type microphones, 
such as condenser type, results from two factors: (1) the variation of 
the diffraction effect and the angle of pickup; and (2) the decrease in 
pressure caused by phase shift which occurs when the direction of the 
sound has a component which is parallel to the plane of the diaphragm. 

Each of these effects is a function of the size of the microphone rela- 
tive to the wavelength of sound. 5 

PRESENTED: October 14, 1949, at the SMPE Convention in Hollywood. 





An extremely small size of the microphone is required in order to 
minimize the variations in directivity with angle. If this effect is to 
be held negligible ( =*= 1 db at 10,000 cycles) the diameter of the micro- 
phone must be approximately no greater than ^ in. The diffraction 
effect is the more important of the two factors described above, where 
the wavelength becomes comparable to the diameter of the dia- 
phragm. This is the effect which makes a difference between the 
"free field" and the "sound pressure" measurement calibration of a 

Fig. 1. Mil microphone system. 




microphone. With these unescapable fundamental principles estab- 
lished, there appears to be only one way in which diffraction distor- 
tion can be reduced to a negligible degree, and that is maintaining an 
extremely small size. Calculated diffraction effects have been 
checked very closely with data obtained from "free field measure- 

The two most commonly used types of housings for pressure oper- 
ated microphones are the cylinder and the sphere. 

When a microphone is built to have the fundamental shape of a 
cylinder, the diffraction effect is such that when sound arrives per- 
pendicular to the end of the cylinder which represents the diaphragm 
(upper part of Fig. 2), the response increases by a factor of approxi- 
mately 8-10 db when the ratio of the diameter to the wavelength of 
sound is unity. At frequencies above this point, a series of variations 
by this same amount occurs at intervals. In the case of a sphere 








2" DIA 
2" DIA. 








" i 

' , 




^ \ 

180 V, 

^ ^^" ^^ 












I50 ,80 



















^ "Z 











.50^ t 

32 0.06 O.I 0.2 0.6 1 2 6 10 
542 1626 2710 5420 16,258 27,096 54,192 162,576 
271 813 "55 2710 8,129 13,548 27,096 81,288 
135 406 77 1355 4,064 6,774 13,548 40,644 

Fig. 2. Effect of diffraction. 




(lower part of Fig. 2), the output perpendicular to the axis is uniform 
at a value equal to that for a ratio of the diameter to the wavelength 
of unity. The cylinder shows a larger variation than the sphere 
which has a minimum variation of 12 db when the ratio of diameter 
of wavelength to sound is greater than 1.5. These data indicate that 
while it might be technically possible to equalize the on-axis response 
to be uniform, a variation of not less than 12 db will be encountered 
for the various angles of pickup for this ratio. 6 - 7 


Having established the fact that miniature size is a necessary re- 
quirement to reduce this form of distortion, effort has been directed to 
design and produce a commercial microphone meeting the require- 
ments of size along with sufficient output to maintain the necessary 
signal-to-noise ratio. Figure 3 shows a cross-section diagram of the 







Fig. 3. Cross section of 2 IB microphone. 

essential portions of this miniature microphone. They consist of a 
diaphragm and an electrode or backplate in close proximity. The 
backplate and diaphragm being closely spaced, constitute an electri- 
cal capacitance which varies with microscopic deflections of the dia- 
phragm caused by pressure variations in the sound wave. The back- 
plate or center terminal is polarized with respect to the diaphragm 
through an extremely high resistance so that a fixed charge accumu- 
lates on the center terminal. As the sound pressure actuates the 
diaphragm, the capacity of the microphone varies and a corresponding 
change in voltage between the center terminals and diaphragm exists. 
The resulting signal is applied to the grid of the vacuum tube which 
follows. The surface of the diaphragm facing the center terminal is 




formed of insulating material, eliminating the problem of electrical 
breakdown between these parts. 

The microphone base encloses a 6AU6 Miniature Vacuum Tube 
whose function is to translate the change in voltage generated by the 
microphone across an extremely high impedance to a nearly equal volt- 
age across an impedance of 1200 ohms so that the signal can be faith- 
fully transmitted over lengths of cables to subsequent apparatus. 
The circuit of the impedance transferring tube in the base is shown in 
Fig. 4. The microphone backplate receives its polarization through 
the elevation of cathode voltage above ground potential. It is a 
property of the cathode follower circuit that its input impedance is 
extremely high whereas its output impedance is relatively low. 8 Also, 
the effect of any capacity connected between cathode and grid is 



Fig. 4. 150A microphone base. 

greatly reduced by the cathode follower action. Connections of the 
inner shield in the microphone permit it to be separated by a short 
distance from the vacuum tube. The extension between vacuum 
tube and microphone is intended to achieve the fullest advantage of 
the miniature size of the microphone without providing any additional 
obstacle size and its associated distortion. The microphone base con- 
tains only the cathode follower vacuum tube. All other components 
associated with it are located at the far end of the interconnecting 
cable. This is done to maintain the smallest possible base size. 

The dimensions of the microphone are .6 in. in diameter and .4 in. 
thick. It weighs approximately 6 grams. A small circumferential 
sound entrance channel, 20 mils thick, is located on the side of the 
microphone. This aids in maintaining an omnidirectional charac- 
teristic and provides protection against mechanical damage to the 
diaphragm. The microphone is mounted on a base which is 8% Q in. 


long and has a diameter of 1 J^ in. at the bottom which has a Cannon 
P-8 Plug. The top of the base is l %& in. in diameter. Since the top 
of the base is smaller in diameter than the microphone, it does not 
contribute any added obstacle interference. 

The microphone and its base are connected to the power supply by 
means of a cable (up to 400 ft in length) whose construction is shown 
in Fig. 5. Signal transmission through this cable is equivalent to the 
use of coaxial cable. The three inner conductors which carry the 
signal have high capacity between them in the circuit. The outer 
conductors carry the other tube functions all of which are at ground 
potential for the signal. By this construction, the outer conductors 
serve to shield the inner signal conductor. 


The power supply cabinet houses the components associated with 
the vacuum tube in the base as well as the output matching trans- 
former, and the plate and filament power. The circuit of the power 
supply is shown in Fig. 6. Rectifiers are the selenium dry disk type 
and supply the necessary plate and screen voltages along with the 
heater current for the 6AU6 Cathode Follower Tube. While the 
cathode follower has a low output impedance, it must work into a 
load of relatively high impedances in order that it function properly. 
In this case, a load of 70,000 ohms is used. The matching trans- 
former is then used to transform this impedance down to conven- 
tional values used in pre-amplifiers. 


In the past a reduction in size of the microphone has lowered the 
sensitivity to such an extent that an adequate signal-to-noise ratio 
could not be maintained. In this microphone the sensitivity is, in 
spite of its size, sufficient to give 48 db below 1 milliwatt for a sound 
field of 10 dynes/sq cm at the output of the transformer. The open 
circuit voltage of the microphone is 50 db below a reference of 1 
volt/dyne/sq cm. The electrical capacitance of the microphone is 
approximately 6 micromicrofarads. The diaphragm is 0.5 in. in 
diameter and is clamped on the edge. Since it is of laminated con- 
struction and part is composed of insulating material, its resonance is 
not determined by tension. 

The frequency response of the microphone, including all apparatus 
up to the output of the matching transformer, is within 1 db from 40 




cycles up to 12,000 cycles. Because of the simplicity of the mechani- 
cal and acoustical elements in the microphone, a very high degree of 
uniformity can be maintained between production units. This uni- 
formity can be observed by placing two microphones immediately 
adjacent to each other. By connecting their outputs so that they are 
out of phase, a sound source will be canceled by a factor of approxi- 




Fig. 5. Cross section of microphone cable. 

152 A OR 
1-iiA CABLE 

Fig. 6. P-518 power supply. 


mately 30 db on a VU meter reading basis. This degree of cancella- 
tion can be achieved by virtue of the fact that since the diameter of 
the microphone is .6 in. the half-wavelength frequency is approxi- 
mately 13,000 cycles/sec. 

In this design the response of the condenser microphone extends 
to very low frequencies (approximately 3 db down at 10 cycles). In 
some cases, it is desirable to eliminate rumblings at low frequencies 
caused by the extraneous noise of traffic and ventilators. A switch 
is provided to reduce the low frequency response at a rate of 6 db per 
octave being down approximately 6 db at 20, 40, and 120 cycles. 


The small physical size of the microphone permits it to approach 
more closely the pattern of an ideal omnidirectional microphone at 
the very high frequencies. This characteristic is a material advan- 
tage where orchestra and solo channels are recorded separately and 
then later combined for proper balance. Directional or semi-direc- 
tional microphones 9 provide less suppression at high frequencies than 
at low frequencies, and when the tracks are combined in recording, 
the high frequency content of the orchestra in the solo channel is dis- 
torted in characteristics and phase, and undesirable effects are ob- 
tained in the release product. 

Use of the miniature condenser microphone for the scoring of music 
in motion pictures has indicated that the usual microphone placement 
practice can be modified. Instead of using several microphones 
placed near individual instruments, one miniature condenser micro- 
phone can be placed so as to obtain a good balance on the over-all 
orchestra. Setups have been made with all of the commonly used 
groups of instruments such as for main title, background scoring, 
small dance band routines, and full symphonic numbers. A con- 
siderable saving in time has resulted from this technique as the time 
required to place and balance multiple microphone setups is elimi- 
nated. The practice to date indicates that the single condenser 
microphone can be placed in a position directly above the conductor's 
head (Fig. 7) . The conductor places the instruments so that the re- 
quired balance is obtained. The sound mixer then has the primary 
function of maintaining the proper modulation on the film whereas 
the conductor is responsible for balance. Where more than one chan- 
nel is used such as in the cases of prescoring with voice, a chestplate 
type of microphone has been used so as to provide the necessary isola- 




tion between the orchestra and solo channels (Fig. 8). This is ac- 
complished by placing the soloist very close to the director. Ap- 
proximately 10 db of suppression on the orchestra is obtained by the 
use of this chestplate since the soloist is permitted to operate ex- 
tremely close to the microphone without the undesirable effects usu- 
ally associated with close operation (Fig. 9). Some of the advantages 
resulting from this technique are: 

(1) The soloist is in a position which closely parallels that of con- 
cert work, and in so being hears the full volume of the orchestra and 
has the maximum benefit of the conductor's direction. 

(2) The orchestra pickup in the solo channel has a frequency re- 
sponse which is equal to the orchestra channel except for volume. 

(3) Re-recording for release is simplified since the phase of the 
orchestra in the solo channel is similar to that in the orchestra chan- 
nel. This permits overlapping without undesirable effects. One of 
these undesirable effects results from the fact that when directional 
microphones are used to provide the necessary suppression between 
orchestra and solo channels, the suppressed frequency response is 

Fig. 7. Typical microphone position for recording orchestra. 


such that the high frequency content is suppressed considerably less 
than the low frequency content. In most cases this suppression 
amounts up to 10 to 15 db between 1000 and 8000 cycles. 

When ribbon microphones or various combinations of this principle 
are used, the phase shift between the particle velocity and the sound 
pressure causes a change in low frequency characteristics which is a 
function of distance from the source. At a distance of 2 ft the increase 
in response is approximately 2 db but at a 6-in. distance the increase 
is 10 db. In the case of the miniature condenser microphone, this 
phase shift does not occur and consequently the microphone can be 
used at distances extremely close to the source without producing this 
false bass. 

In sound motion picture recording the microphone is one of the 
very important links in production since it involves the final quality 
and the various compromises that must be made by set designers, 
cameramen, and sound departments in order that the microphone be 
placed in its optimum position with a minimum of delays to the pro- 
duction company. Elaborate booms have been provided for both 
"panning" and "tilting" the microphone. Panning is required in 
order that fixed distances be maintained between the actor and micro- 
phone for best and uniform quality. Tilting has been required be- 
cause the microphone has a varying high frequency response, depend- 
ing upon the angle between the face of the diaphragm and the actor. 
The size of the boom has been dictated by the weight of the microphone 
and the mechanism required for panning and tilting. An extremely 
lightweight microphone reduces the demand for heavy construction 
and counterweighting. The elimination of tilt further reduces the 
complexity of controls on the boom. Use of condenser microphones 
in general, over a period of years, has indicated that they are less 
susceptible to wind disturbances than other types of microphones 
which have been used in motion picture production. A very small 
wind screen consisting of four layers of voile is provided as an accessory 
to the microphone. The attenuation of this material does not exceed 
1 db at 10,000 cycles and provides up to 15 db suppression of wind 


This paper has stressed the improvements in sound quality and 
directional effects caused by the small size of the microphone. All of 
the features discussed contribute to produce a readily discernible 


Fig. 8. 155A chestplate microphone. 

Fig. 9. Relative position of soloist to conductor and orchestra. 


improvement and a distinct step forward in the quality which can be 
obtained in sound reproduction. Experimental use over several 
months has indicated that under conditions varying from Carnegie 
Hall broadcasts to stage, motion picture, and television, obvious im- 
provements in sound quality have been attained. 

NOTE: In addition to the paper, a demonstration at the Academy Award 
Theatre of various types of recording, which included main title, background, com- 
bination of orchestra and solo channels, was presented. It was derived from 
material made by Mr. Alfred Newman of the Music Department at 20th Cen- 
tury-Fox Studios and the co-operation of their Sound Department. 


(1) Amos E. Dolbear, "On telephone systems," /. Frank. Inst., pp. 20-23; 
January, 1886. 

(2) E. C. Wente, "The sensitivity and precision of the electrostatic transmitter 
for measuring sound intensities," Phys. Rev., vol. 19, p. 498; May, 1922. 

(3) H. C. Harrison and P. B. Flanders, "An efficient miniature condenser 
microphone system," Bell Sys. Tech. J., p. 451; July, 1932. 

(4) F. L. Hopper, "Characteristics of modern microphones for sound recording, 
Jour. SMPE, vol. 33, pp. 278-288; September, 1939. 

(5) S. Ballantine, "Effects of diffraction around the microphone in sound 
measurements," Phys. Rev., vol. 32, p. 988; December, 1928. 

(6) R. W. Marshall and E. F. Ramonow, "A non-directional microphone," 
Bell Sys. Tech. J., vol. 16, p. 45; July, 1936. 

(7) H. F. Olson, Elements of Acoustical Engineering, Van Nostrand, New York, 
1947, p. 18. 

(8) J. G. Frayne and H. Wolfe, Elements of Sound Recording, Wiley, New York, 
1949, chap. 27, p. 544. 

(9) R. N. Marshall and W. R. Harry, "A cardioid directional microphone," 
Jour. SMPE, vol. 33, p. 260; September, 1939. 

Supplementary Magnetic Facilities 
For Photographic Sound Systems 


Summary To facilitate the introduction of magnetic recording on 35-mm 
film, modifications have been engineered for adapting photographic re- 
cording and reproducing systems so that they may be used alternatively 
for either photographic or magnetic recording. Existing transmission 
systems employed in photographic recording have been modified to include 
a bias-erase oscillator. Magnetic heads have been added to the film re- 
corders in such a manner still to permit the use of the recorder for photo- 
graphic recording. Re-recorders have been modified for magnetic only 
or for alternative magnetic or photographic reproduction. The existing 
photocell amplifier has been modified for the dual reproducing facilities. 

To FACILITATE the introduction of synchronized magnetic re- 
cording into sound motion picture studios, it is highly desirable 
during the transition stage from photographic to magnetic to utilize 
as much as possible of the existing sound recording and reproducing 
equipment with a minimum of modifications. This serves the dual 
purpose of lessening the economic burden of the change-over and at 
the same tune expanding the over-all recording facilities of the 
studios. As magnetic recording takes a firmer hold in sound motion 
picture production, emphasis will undoubtedly shift toward using 
facilities intended primarily for magnetic recording; but in the in- 
terim period the use of the dual-purpose equipment will meet with 
much favor. 

At the time of writing this paper, the use of magnetic recording in 
the motion picture studios has not progressed much beyond the stage 
of using the medium for original recording. One of the dual equipped 
recorders described below may be employed for this purpose. The 
accepted takes are usually transferred from magnetic to photographic 
tracks of either the density or the area type, and the usual photo- 
graphic re-recording and cutting techniques are then followed from 
this point. In transferring to photographic film, the "magnetic only" 
re-recording machine described below is preferred, and the photo- 
graphic recording may, as in many small studios, be done on the dual- 

PRESENTED: October 14, 1949, at the SMPE Convention in Hollywood. 



purpose recording machine. With the equipment described below 
either 200-mil push-pull area or density, or 100-mil single density 
or area may be recorded. Provision is also made for transferring the 
magnetic to either a direct-positive area or density track. 

The design details of the components and the operating charac- 
teristics of the magnetic systems described below will not be re- 
viewed in this paper as they have been thoroughly discussed in an 
earlier paper. 2 It will be noted that although the high-frequency 
erase facilities previously described are provided in the systems de- 
scribed below, the current practice in the Hollywood studios is to use a 
bulk 60-cycle eraser rather than the erase head mounted in the 
recorder. The use of low-frequency pre- and post-equalization has 
become well established in the industry, but the use of high-frequency 
pre- and post-equalization remains optional at this time. 


The recording and reproduction from magnetic film are being 
accomplished by two general methods with relation to the film path 
encountered in existing Western Electric film recorders of the RA-1231 
type. The first consists of placing the magnetic head so that it bears 
on the overhanging portion of the film at the scanning drum in a 
manner analogous to that used in photographic reproduction. This 
is frequently referred to as the preferred position, since flutter and 
amplitude modulation are minimum at this point. The second 
method is that of placing the head in an open film path, such as be- 
tween the recording drum and one of the compliant filter rollers, and 
this method is frequently referred to as loop scanning. The per- 
formance under this condition is dependent upon several design fac- 
tors, but in general is not as good as the drum position with respect 
to high-frequency flutter rates. Typical flutter values for the drum 
position in either recording or reproduction are .06% to .08% total 
flutter, most of which is found in the higher flutter rates with low- 
rate components very small, the over-all performance being com- 
parable to photographic recording. Flutter values for the loop 
position are substantially the same at all lower flutter rates, but are 
somewhat greater at the higher rates, particularly at the 96-cycle 
sprocket hole rate. The amount is somewhat dependent upon the 
relationship between sprocket pitch and that of the film, but for 
normal operating conditions is well below the accepted threshold 
limit for perceptibility. 2 It has been shown 3 that the fundamental 


relationship between flutter rate and its minimum perceptibility will 
permit much larger values at high rates. For example, flutter at 
96-cycle rate may be nearly 10 times that at rates between 5 and 10 
cycles for equivalent perceptibility. 

The addition of magnetic recording and monitoring facilities to the 
RA-1231 film recorder is shown by Fig. 1. The magnetic head at the 
drum position is shown as the light-colored circular object located 
in front of the recording drum and bearing upon the overhanging 
edge of the film which has the magnetic coating on the inside surface. 
An identical magnetic head used for monitoring may be seen at the 
right just above the film and located in the film loop between the 
recording drum and the lower right-hand filter roller. Both heads 
are adjustable for azimuth and track position and the recording head 
mounting is pivoted so that it is pressed against the film with a con- 
trolled pressure at all times to compensate automatically for factors 
such as film curl. This arrangement maintains proper contact at all 
times between the surfaces of the head and the film coating, and its 
motion is well damped to maintain stability and damp out small 
transient motions such as those produced by a splice passing the head. 
The recording-head mounting is also provided with a device for re- 
tracting the head and holding it away from the film to prevent con- 
tact during photographic recording, particularly in the case where 
the modulator is moved toward the front of the machine to record 
on the outside edge of the film for producing direct-positive sound 
tracks without the necessity of reversing the direction of motion of the 
film in the recorder. 

Good contact is maintained between the film and the monitor head 
by virtue of the uniform and constant tension in the film path be- 
tween the two sprockets. 

A magnetic erase head is also available for application to this 
recorder, although it is not shown in Fig. 1. When required, this 
head is mounted in a manner similar to the monitor head and is 
located to the right of the film in the path between the left-hand 
sprocket and the left-hand filter roller so that the film may be erased 
ahead of the recording point. 

For magnetic recording, the film magazine is replaced by a reel 
adapter mounted in the same manner as a magazine. It permits the 
use of either the 10-in. sheet-metal reel having a 2-in. hub, or the 
preferred reel having an 11-in. diameter with a 4-in. hub. All mag- 
netic conversion parts are available in kit form and may be applied 










to any of the Western Electric RA-1231 type of film recorders without 
impairing in any way their usefulness as photographic recorders. 
Thus it is possible to have alternative magnetic variable-density 
recording or alternative magnetic and variable-area recording on 
any of these film recorders. 


Magnetic reproduction may be roughly divided into three classi- 
fications: immediate playback following recording (which for con- 
venience should be done on the same machine making the recording), 
magnetic re-recording operations requiring adaptation of existing 
re-recording machines and magnetic reproduction in theater-type 
reproducers for use in the studio review room. 

The immediate playback is available by using the monitor facility 
of the RA-1231 Recorders as described above, or in certain cases where 
high-quality performance is required, the recording head may be 
electrically reconnected and used for reproduction. 

For the re-recording of magnetic tracks, two modifications have 
been made available for the RA-1251 type of re-recording machines 
recently described, 4 in which both the drum and loop scanning posi- 
tions may be utilized. The first of these uses the preferred drum 
position as shown by Fig. 2, which is a close-up of the center section 
of the machine. All of the facilities for photographic reproduction 
are removed and a magnetic head is placed within the film drum 
contacting the over-hanging position of the film in the same manner 
as in the recorder. The magnetic head is pivoted and mounted in 
the same manner to insure adequate contact between the head and 
film under all conditions of operation. Similar adjustments for head 
orientation, azimuth, track position and pressure are provided. 
The only changes required in the film path or machine facilities are 
those directly related circuit functions pertaining to the equipment 
removed and the transmission facilities which are described later. 
The second modification consisting of placing the magnetic head in 
the loop position is shown in Fig. 3. This permits the retention of 
the standard photographic reproducing facilities so that the machine 
is available for either magnetic or photographic operation. The 
magnetic head is contained within the small rectangular unit mounted 
between the impedance drum and lower filter roller as shown by Fig. 3. 
This unit also contains a fixed idler roller so that for photographic 




operation, the film is threaded over this roller instead of through 
the slot as shown in the figure. The change in film-path angle be- 
tween these two methods of threading also serves to compensate for 
the added friction of the film over the magnetic head so that the two 
filter rollers operate in substantially the same positions with both 
types of film. This modification requires little change in the basic 
machine, except for the addition of the head assembly as shown 

Fig. 2. Close-up of RA-1251-B Re-recorder equipped for magnetic reproduc- 
ing (or recording) at the drum position for optimum flutter performance. 

which replaces a fixed roller. The magnetic head is terminated in a 
plug on the rear of the removable mechanism section and provision 
has been made so that the photocell amplifier may be utilized as a 
magnetic head amplifier to provide an over-all flat frequency re- 
sponse, as described later. 
Both of these modifications are available in kit form for application 




to existing equipment, and in the event that either of these magnetic 
modifications is desired for recording purposes, an erase head unit 
has been designed to be mounted in much the same manner as de- 
scribed for the recorder. Figure 3 shows this erase head assembly 
mounted in the film loop between the upper sprocket and the first 
filter roller in combination with the photographic and magnetic 

Fig. 3. Close-up of RA-1251-B Re-recorder equipped for both photographic 
and magnetic reproduction and erasure, with the film shown threaded for mag- 
netic operation passing over the magnetic head under the cover located below and 
to the right of the impedance drum. 

modification. A knob on this unit retracts the head from contact 
with the film during photographic operation to avoid film damage, 
but does not in any way affect the other performance characteristics 
of the film pulling mechanism. A similar assembly is available for 
the other modification employing drum scanning, but it is of course 
located on the other side of the film, since the coating is on the other 


side in this type of operation. In this case the magnetic head is 
fixed, without the retractable feature, since photographic film is not 


A new theater reproducer, which is primarily intended for wide- 
track photographic reproduction in studio review rooms, has been 
described recently. 5 Figure 2 of that previous paper shows there- 
producer equipped with a magnetic head located in the loop position 
directly below the scanning drum. Since this machine is designed 
for both photographic and magnetic operation, the magnetic head is 
also retractable so that it does not contact photographic film, to avoid 
scratching. The flutter performance is essentially the same for 
photographic and magnetic except for the higher values at higher 
rates as discussed above. 


The transmission facilities of both the portable (300 type) 6 and 
deluxe (400 type) 7 recording systems have been adapted for use with 
the magnetic recording machines described above. As in the case of 
the recorders the modifications have been made in such a way that 
the photographic facilities are retained and change-over between the 
two types of operation is readily made. The modification parts 
have also been provided in kit form to permit conversion in the field. 


Although compactness was one of the primary features of the 
"300" system design, it has been possible to add the magnetic and 
magnetic-photographic change-over facilities without adding to the 
number of component units or increasing the over-all size of any of the 
units. All changes have been confined to the recorder, including the 
photocell monitor assembly mounted within it, and the main trans- 
mission unit. The adaptation has been applied to both the variable- 
density and the variable-area versions of these units. 

The modifications of the main transmission unit for alternative 
variable-density photographic or magnetic recording are described 

High-Frequency Bias. A high-frequency bias signal superimposed 
on the speech signal is normally used in high-quality magnetic re- 
cording systems to minimize distortion caused by the nonlinear 


characteristics of the magnetic medium. 1 In this system it has been 
possible, by relatively minor circuit modifications, to utilize the 30-kc 
carrier frequency oscillator associated with the noise-reduction cir- 
cuit as a source of high-frequency bias signal. 

In the modification, the signal frequency is changed from 30 to 60 
kc to obtain a greater spread be'tween speech and carrier for mini- 
mizing intermodulation components. This change does not affect 
the performance when operating as a noise-reduction circuit. 

To prevent modulation of the carrier by the speech signal (as is 
done in normal noise-reduction operation) the oscillator is discon- 
nected from the rectified audio signal under magnetic operation. 

In the noise-reduction output circuit, the output transformer is 
replaced with one of equivalent performance except for an added 
600-ohm output winding to operate into the recording magnetic head. 
Also a means of metering the high-frequency signal using the D-C 
Noise Reduction Bias Meter is provided. The oscillator circuit is 
capable of supplying over 30 milliamperes to the output circuit, 
which is more than ample for optimum recording conditions. 

Frequency-Response Characteristic. The low-pass filter normally 
used in 16-mm recording has been modified to provide a relatively 
sharp cutoff at 9,000 cycles per sec for magnetic recording. This 
prevents the high-frequency bias from feeding back into the amplifier 
output circuit where it would affect the volume indicator or meter 
reading and possibly introduce intermodulation components. 

Power Supply to Monitor Amplifier. The heater supply to the 
associated monitor amplifier in the recorder has been connected 
to the 12-v d-c line instead of the 6.3-v a-c line. This retains the 
monitor system hum level at a satisfactorily low level, whereas it 
would otherwise be emphasized under magnetic operation due to the 
6 db per octave slope of the reproducing system frequency-response 
characteristic. A ballast lamp is added for current regulation under 
the conditions of variable alternating-current supply voltage and 
variable current drain to the recorder lamp. 

The variable-area version of the main transmission unit has not 
previously been described in the JOURNAL. It differs from the vari- 
able-density unit principally in that peak-chopping has been pro- 
vided, the light-valve equalizer has been modified to provide charac- 
teristics complementary to those of the variable-area light valves 
and the noise-reduction filter has been adjusted for optimum variable- 
area attack and release times. 


The magnetic facilities are essentially the same for the variable- 
density and variable-area versions of the main transmission unit, 
the peak-chopper circuit in the latter version being disabled when 
recording magnetically in order to utilize fully the gradual overload 
characteristic of the magnetic medium. 

By furnishing the modification parts in the form of wired sub- 
assemblies, the change-over can be made economically in the field 
with a minimum disturbance to the original components and wiring. 

Monitor Amplifier in Recorder. Alternative magnetic and photo- 
graphic film monitoring is accomplished by installing in the rear 
compartment of the recorder a new monitor amplifier. This amplifier 
has a photocell input for photographic operation from the modulated 
light beam and a low-impedance transformer input for operation from 
the magnetic monitoring head. An internal transfer switch connects 
the desired input circuit and also selects between two feed-back cir- 
cuits around the first two stages; one provides a flat characteristic 
for photographic and the other a 6 db per octave slope for magnetic 

With magnetic operation a signal-to-noise ratio better than 50 db 
can be obtained from an over-all flat reproducing system. This 
value is considered adequate for portable production applications and 
therefore pre- and post-equalization has not been added. 


In the deluxe, "400," photographic recording system, such as used 
in major studios for production, scoring and re-recording, the various 
electronic components are separately packaged units adaptable for 
mounting in standard equipment cabinets or custom-built consoles. 
These transmission assemblies normally operate in conjunction with 
the deluxe variable-density or variable-area recorder and the auto- 
matic recorder control cabinet. 

Modification of these systems for alternative magnetic and photo- 
graphic operation has been accomplished by providing modifications 
of two of the existing system components and adding two new com- 
ponents as described below. 

Photocell Monitor Amplifier. This amplifier, mounted in the re- 
corder control unit, normally operates from a photocell mesh circuit 
and provides a flat over-all frequency response for monitoring at an 
output level of approximately dbm. By adding an input trans- 
former, a magnetic reproducing characteristic of 6 db per octave 


slope and a switch for transferring between the two circuit conditions, 
the amplifier can be made to function for both types of recording. 
The output attenuator can be adjusted as required to provide direct 
and film-monitor balance in both cases. This modified amplifier is 
also used in the modified re-recorders described above to provide a 
flat over-all response characteristic when re-recording or reproducing 
from magnetic film. 

Recorder Control. A feature of the standard photographic recorder 
control unit is the provision for automatic switching, in proper time 
sequence, of the recorder speech input, noise-reduction bias, recorder 
lamp, motor and other miscellaneous circuits at the beginning and end 
of each take from a single start-stop push button. This is accom- 
plished by a positively driven multiple-cam assembly operating on a 
series of individually adjustable micro-switches. Protective inter- 
locking features are also included. 

In the magnetic modification both the manual and automatic 
operating features are retained for both types of operation. By 
operating two transfer switches within the unit the following circuit 
changes are made: 

(a) Speech Off-On switch is transferred from the light valve to 
the magnetic head circuit. 

(b) Power to the light meter and noise reduction unit is trans- 
ferred to the bias-erase oscillator which is one of the added units 
described below. 

(c) The meter used for measuring exposure is transferred to meas- 
ure bias and erase current. 

In addition, several items are added on completely wired sub- 
assemblies including bias and erase controls, high-frequency suppres- 
sion filter (to prevent the bias current from affecting the volume 
indicator reading) and a high-frequency post-equalizer. 

The latter is used in the monitor circuit to complement a high- 
frequency pre-equalizer added in the recording circuit as described 

Bias-Erase Oscillator. This new unit is added as a part of the 
system modification. It is a single-stage, push-pull oscillator em- 
ploying two 6L6 Vacuum Tubes. In the output circuit a tuned filter 
provides an output signal having approximately 0.1% even har- 
monics. This prevents the introduction of objectionable noise and 
distortion due to an unsymmetrical bias or erase signal. Since the 
output control potentiometer is in the common cathode circuit, the 


total plate drain varies with the signal level. By this means the 
plate drain is reduced from approximately 75 milliamperes for bias 
plus erase, to 30 milliamperes for bias only. The oscillator is capable 
of providing a maximum load current of 275 milliamperes at an output 
voltage of approximately 65 volts. 

Pre-Equalizers and Recording Circuit. The low-frequency pre- 
equalizer has an 8-db peak at 60 cycles per sec. The high-frequency 
pre-equalizer is the conventional one used for many years in many 
Hollywood studios which has a 12-db rise at the high frequencies. 
The use of these two equalizers increases the effective signal-to-noise 
from approximately 50 db to approximately 58 db. Complementary 
low-frequency post-equalization in the monitoring circuit is obtained 
by properly selecting the low-frequency shelf-point for the 6 db per 
octave reproducing equalizer. A high-frequency post-equalizer is 
provided in the monitor circuit of the recorder control and mixer 

A general purpose amplifier such as is used as a mixer and booster 
amplifier is provided to operate into the magnetic recorder head. 
This is preferable to utilizing the standard limiting amplifier because 
of its lower noise level and more gradual overload characteristic. 
However, the limiting amplifier is retained to operate the volume in- 
dicator and direct monitor circuits in the normal fashion, which greatly 
simplifies the circuit changes required when transferring between the 
two types of operation. For magnetic operation the limiting action 
is disabled and the peak-chopping level is set to its maximum value in 
order that it will affect only the occasional, short-duration overloaded 

The added components, consisting of bias-erase oscillator, low- 
and high-frequency pre-equalizers and magnetic recording amplifier, 
can either be mounted in the console or cabinet of a particular system 
or may be assembled in a small portable cabinet with patching facili- 
ties for connecting to any system as required. 

Since the modifications made in the standard system include trans- 
fer of power from the light meter and noise-reduction unit, no addi- 
tional power unit is required when connecting the special magnetic 

The magnetic modification facilities described herein have made 
available to the industry an immediate and economical means of 
utilizing existing studio equipment for both experimental and pro- 
duction magnetic recording. 


Experimental investigations will allow the studios to familiarize 
themselves with this new medium and to assess its technical and 
economic value as applied to their particular production require- 

Production experience with this modified equipment will also form 
the basis for the establishment of design requirements for the equip- 
ment which will eventually be made available for the magnetic re- 
cording of sound for motion pictures without the encumbrances of 
photographic recording facilities. 


(1) J. G. Frayne and H. Wolfe, "Magnetic recording in motion picture tech- 
niques," Jour. SMPE, vol. 53, pp. 217-234; September, 1949. 

(2) Report of SMPE Committee on Sound, "Proposed standard specifications 
for flutter or wow as related to sound records," Jour. SMPE, vol. 49, pp. 147-159; 
August, 1947. 

(3) J. G. Frayne and H. Wolfe, Elements of Sound Recording, chapter 22, Wiley, 
New York; June, 1949. 

(4) W. C. Miller and G. R. Crane, "Modern film rerecording equipment," Jour. 
SMPE, vol. 51, pp. 399-417; October, 1948. 

(5) G. R. Crane, "Theater reproducer for double width push-pull operation," 
Jour. SMPE, vol. 52, pp. 657-661; June, 1949. 

(6) E. W. Templin, "35 mm and 16 mm portable recording system," Jour. 
SMPE, vol. 53, pp. 159-182; August, 1949. 

(7) F. L. Hopper and E. W. Templin, "New deluxe sound recording equipment 
and its system applications," presented October 27, 1948; SMPE Convention in 

Sprocketless Synchronous 
Magnetic Tape 


Summary Advantages inherent in the normally thin, narrow magnetic 
tape may be realized with a new method of obtaining true synchronism 
with film for motion pictures. Adaptability of the system to dubbing and 
post-synchronization is described. For television sound, it offers wide- 
frequency response and long playing time. 


_/\. nevertheless been a boon to the motion picture engineer since 
Edison's day. All motion picture equipment has been built around 
the sprocket hole; and then, of necessity, very thorough effort has 
been made to eliminate sprocket-hole flutter. Quite naturally, there- 
fore, when magnetic recording showed its great possibilities, the first 
thought was to coat normal sprocket film base with magnetic coating, 
substitute magnetic heads for light valves and phototubes and carry 
on largely with the normal equipment. A very complete discussion of 
this approach has been given in the JOURNAL. l There are also several 
other JOURNAL articles of interest. 2 Concentrated efforts have also 
been made to place a narrow strip of magnetic coating on normal pho- 
tographic stock so that the sound could be recorded on this strip, 
while the rest of the film carried the normal picture. This appears to 
be the ultimate system for prints. 

Narrow magnetic tape has come to the fore. It is only .002 in. 
thick and has definitely established its position in sound reproduc- 
tion. Its very thinness may hold some of the secret of its success; 
and, of course, the quarter-inch width makes for lighter and more com- 
pact reels. The application of this thin tape to synchronized opera- 
tion has been so well developed that it is now successful in a practical 
manner. Also, during this development other advantages of this 
method of operation have been discovered, and these will be described. 

This new system is independent of slippage and possible stretch of 
the sprocketless tape ; these exist and perhaps always will. Further- 
more, even if the tape drive were perfect, that of the motion picture 
film is not; it depends on the frequency of the power driving the cam- 

PRESENTED: October 14, 1949, at the SMPE Convention in Hollywood. 


era motor. This may vary from instant to instant and day to day. 
What is really essential is not absolute speed regulation, but relative 
uniformity of speed between tape and film. As some form of regis- 
tered control is absolutely essential, this system without actual per- 
forations uses what might be termed "magnetic sprocket holes." 
They actually are magnetic recording of the pulses of the power driv- 
ing the camera at the same time that the sound is being made. 
Normally this is 60-cycle power, so on the tape this power frequency 
is recorded simultaneously with the sound that is to accompany the 
pictures being shot. This power-frequency recording will then be 
used as a control to keep projector and tape in step for subsequent re- 

One immediate advantage that this method of registering control 
gives is that it is unnecessary to work with any kind of speed control 
in the tape recording other than the normally smooth advance of the 
tape accomplished with the aid of the synchronous motor-driven cap- 
stan of the tape machine. A good recording is assured with normally 
engineered tape equipment. 

The sound recording on the tape is accomplished with normal mag- 
netic heads, shown as the center three of the five of Fig. 1. These 
three are the "Erase" to clear the tape of any previous recording, 
then the "Record" which puts down the normal sound recording, and 
finally the "Playback" which gives the immediate playback to monitor 
the recording and then the final playback for recording. It is to be 
noted that all of these heads are magnetic toroids with the gaps in the 
toroids extending vertically across the tape. As the magnetism var- 
ies in these heads the tape will be correspondingly magnetized longi- 
tudinally, i.e., in the direction of travel of the tape. 

Now come the two control heads. The one on the left, which the 
tape traveling to the right first hits, is the "Playback Synchronizer." 
The one on the right is the "Record Synchronizer." It will be noted 
that, although these are also toroids, they are oriented at right angles 
to the other three; and this places their gaps in line with the tape 
movement. The magnetic effect in the tape with which they are 
concerned will therefore be up and down or vertical to the tape move- 
ment. This reorientation makes it possible to put the control signals 
on the tape without interfering with the normal sound recording, for 
two reasons : first, the control signal is made in a very narrow band in 
the center of the tape; and second, with its magnetism at right angles 
to the normal sound magnetism, the effect is definitely minimized. 
Actually it is possible to adjust the relative angles of the heads to ob- 




tain this minimum; and normal sound level will be well above 50 db 
stronger than the remanent control signal which is picked up by the 
normal " Play back" head. 

A large portion of motion picture work is made "double system," 
one equipment for the picture, the other for the sound ; so here there is 
virtually no additional equipment necessary for tape sound instead of 
film sound. As a matter of fact, the tape sound equipment is lighter 

Fig. 1. 

Arrangement of 
magnetic heads. 

Fig. 2. Field unit, with recorder and microphone pre-amplifier. 

than most sound film equipment. Furthermore, and here is'the next 
big advantage of the tape, enough for an hour's steady recording can 
be readily accommodated in the tape reel. 

On location the sound man sets up his portable equipment (see Fig. 
2) about twenty-five feet away from the camera. A cable goes to the 
pre-amplifiers from which three microphones may be controlled with 
high-level mixing, earphones and VU meter. A preliminary run is 
made on the sound without the camera, and the director is able to call 
for a playback immediately to judge the quality and placement. After 


this first test, it is usual for the sound man to run the tape back to the 
start and then be ready for the next, or real, take. This will erase 
what has been previously recorded. But after this first test, it is 
never wise or necessary to go back and erase any subsequent takes. 
That they are "Out-Takes" is just noted on the log, and the next take 
proceeds. It is of course wise to have blind track made on the scene, 
to have the normal background noises for subsequent splices which 
will then give good uniformity with the normal background noise. 
Also, these "Out-Takes" can often save the day by giving replacement 
material to substitute for some extraneous noise which may have got- 
ten into what would otherwise be a good take. It is well to save ev- 
erything once recording starts. 

Incidentally, it is interesting to note that this synchronizing tech- 
nique lends itself to all magnetic tape recorders of this variety. To 
adapt an instrument for this work, a synchronizing record head and 
associated circuits for recording the sixty cycles are simply added to 
the recorder. The recorder operates in its normal way otherwise. 

Editing and splicing tape are relatively simple ; there is no blooping 
problem. Also no attention need be paid the "magnetic sprocket 
holes." The greatest possible error in splicing them would be one- 
half a cycle which is less than one-quarter of a frame. The law of av- 
erages produces advances half the time and delays the rest, and to date 
no noticeable discrepancy has been discovered in an accumulated 
frame error from this source. Two continuous music sequences may 
be spliced without any noticeable bloop in the sound, although it is 
better for cadence values to make splices in silent portions. 

When the takes have been assembled, they are then ready for dub- 
bing to sound negative film. It is, naturally, best to run through the 
sequences to furnish the sound engineer, who is to do the dubbing to 
film, a chance to determine his dynamic values and frequency equali- 
zation before the film is exposed. The series of sequences is then run 
through as quickly as the series of film reels may be changed. The 
frequency range available from the tape is from 40 to 15,000 cycles, 
so that the dubbing engineer has a great chance to get the maximum 
out of the recording onto the film. Some very successful full orches- 
tral selections have been made by this process. 

To synchronize the tape to the film recorder, two methods are 
available. The first is to take the control signal recordings from the 
tape, amplify them sufficiently to control a thyratron inverter, shown in 
Fig. 3. This inverter gives 250 watts at the exact frequency of the 
control signals on the running tape. If this energy is delivered to the 




film recorder, the sprocket holes of the film in this recorder will be ad- 
vanced in strict accord with the magnetic control on the tape, so that 
perfect synchronism results. 

Another method is to compare the control signals coming off the 
tape with the 60-cycle power then being used to drive the film re- 
corder. This comparison may be made automatically in a bridge ar- 

Fig. 3. Thyratron inverter. 

rangement operating on a vacuum reactance tube as given in Fig. 4. 
This reactance tube in turn controls the frequency of an oscillator 
tube set normally for 60 cycles. This oscillator tube then drives, the 
thyratron inverter previously mentioned, but which is now used to 
drive the synchronous motor on the tape machine. If the signal on 
the tape happens to be exactly in step with the power frequency at 




the moment, the bridge is balanced and the reactance tube does noth- 
ing to the oscillator. But if the tape gets ahead so that the control 
frequency on the tape is ahead of the power source, then the bridge 
becomes unbalanced to the point of slowing down the tape advance, 
until the control signals on the tape are exactly in step with the power 
frequency. The tape is moved ahead if it gets behind by the reverse 
process of course. It is necessary to make this control quite slow so 
that it will cause no "wow" in the sound. It has been found that if this 
correction is at a rate of less than one cycle in two seconds, no speed 
variation can be heard. 

Post-synchronizing offers a very interesting application of tape for 
the sound. In this operation, the sound is recorded on tape with the 
best possible attention to microphone placement. For this work, the 
control signal is not put on the tape at this time as the cameras are not 

Fig. 4. 

in action. Instead, the control signal is put on during the subsequent 
playback amplified to a loudspeaker to which the actors are perform- 
ing for the cameras. So the control "Synchronizing Record" head is 
energized at the tune that the cameras are likewise running in step 
with the 60-cycle power then in use. Thus, the control signal gets on 
the tape in step with the film for the pantomimed action. If long 
shots and close-ups are to be made from the same sound playback at 
different times, the synchronizing control signal is put on the tape dur- 
ing the long and usually continuous shot. Then the automatic con- 
trol is used for the short close-ups to make the tape run in synchro- 
nism with the cameras then being used. 

Tape has a third interesting advantage. With its flexibility, the 
tape may be made to accommodate or correct what often are slight 
errors in the timing of the actors playing to sound. For example, it 
was found that a very important orchestra always had a tendency to 
speed up over its first rendition of a given piece when it was perhaps 

334 R. H. RANGER March 

straining a bit to keep in strict step with the sound; therefore, after 
several attempts, the engineer simply speeded up the tape very 
slightly on the playback for camera action and the net result was per- 

This may be carried a step further in a method that has already 
been tested, namely, a good musician controls the speed of the tape on 
playback and keeps the tape sound in step with that of the orchestra. 
This relieves the orchestra and the conductor of any strain in trying 
to do a double job themselves. So the orchestra and the conductor 
carry on in their normal manner after once having got the start from 
the loudspeaker playback, and the musician at the controls does the 

Now, as for dubbing: in this case the picture is made first and the 
sound subsequently recorded from the actors who are following the 
film pictures as they are projected. Tape now shows a fourth advan- 
tage. First, any number of tries may be made without wasting record- 
ing stock and everyone knows that it is only with extreme dif- 
ficulty that a perfect job can be accomplished, no matter how many 
the tries. But with tape, discrepancies may be corrected. For this 
purpose, loops of tape and picture film are made. As good as possible 
a take of the sound with the picture is made. The control signal is on 
the tape for this sound, and it is in step, supposedly with the film. 
But at certain portions the sound is a trifle late or early. (It is much 
worse if it is early. Distance gives a customary lag.) So the tape 
should be advanced or retarded to correspond. As is noted in Fig. 1, 
the "Synchronizing Playback" head comes first in the head assembly. 
So with the automatic control on, this head will see to it that the tape 
moves forward in step with the film. An indicator on the front of the 
Automatic Synchronizer control indicates Frame Advance or Retard 
from Zero accord, as shown in Fig. 5. But the automatic control may 
be thrown off when desired, and manual adjustment be substituted. 
Then the tape may be moved forward or back an appropriate number 
of frames. The operator tries this in dry runs until he makes the two 
agree very nicely, then he registers the new control once and for all. 
To do this, he goes through the same manual control, but turns on 
just the "Synchronizing Record" head. 

Unpremeditatedly, it has turned out that this construction of 
"Synchronizing Record" head and the type of control signal put down 
on the tape may be overridden very completely by a second applica- 
tion of energy to this same head. So, in effect, a previous control sig- 
nal may be completely replaced with a new one when desired. There- 




fore, as the operator advances or retards the tape by frames, the new 
control signal is registered with the energy turned into this head, and 
thereafter automatically controlled playback will keep to this new 
step. Furthermore, if slight additional correction is necessary the 
process may be repeated. 

Just one precaution: it is necessary when making such manual 
changes to be sure to come back to zero frame deviation at the end of 
the tape loop, so that the next run-through of the loop will start with 
the film. Also, another advantage of the positioning of the ''Synchro- 
nizing Record" in the succession of heads gives the operator his re- 
action time to make the frame change slightly after he has observed 
the necessity for it. The separation between these end heads is 

Fig. 5. Manual or automatic control unit for synchronous tape drive. 

in., so at 15 in. a second, the normal tape speed, he has half a second to 
make the change indicated. The net result is that he can do it very 

Sound for television gives tape a fifth advantage. A double system 
of operation may be used for simultaneous transmission of pictures 
from film and sound from tape. Tape with its top frequency capabil- 
ity of 15,000 cycles offers full range to the FM sound transmission of 
television. Furthermore, it is no problem at all to run a straight half- 
hour of sound from tape, synchronized with the film. 

The methods of starting tape and film together are to be noted. One 
method is to have normal start marks on tape and film, to place the 
film in the projector at this start mark on the film and to use a strip of 
metal foil stuck to the back of the tape at the start mark on the tape 
to start the film projector. The tape is started with a short leader 
ahead of this metal foil. Then when the tape is up to speed, along 

336 R. H. RANGER 

comes this foil strip which makes a brief connection with a contact on 
the tape machine, and this closes a relay which starts the film pro- 
jector. The time that it takes the projector to come up to speed is 
fairly regular, and it can be calibrated once and for all so that repeated 
starts may be made by this process without difficulty. 

Another similar method for making the start also uses a leader on 
the tape ahead of the start mark. This leader has control signals on it 
for all but the last inch of the leader ahead of the start mark. A sensi- 
tive relay in the amplified output of the control signal operates on the 
control signals, and then when the momentary halt in them comes 
along, the sensitive relay drops back on its back contacts and the pro- 
jector relay is set for this condition to get its start, so that then off 
both go together. Subsequent operation of the sensitive relay will 
have no effect on the projector relay, so that any momentary changes 
in the control signal will not affect the operation. 

If at any time it becomes apparent to the operator that the tape 
needs to be advanced or retarded with respect to the film because of 
an error in threading or editing, something which he can become very 
expert in observing, the automatic synchronizing lock may be thrown 
off, and the framing be adjusted by hand smoothly during operation. 

Shows are now being televised in which tape is used alternately 
with live program, with excellent uniformity between the two methods 
of presentation. This makes feasible a new type of program. 

Tape is making real advances in its application to the motion 
picture field. Some of these have been most unexpected, and there 
is every reason to believe that as more people become acquainted 
with its potentialities, other uses will develop. 


(1) J. G. Frayne and H. Wolfe, "Magnetic recording in motion picture tech- 
niques," Jour. SMPE, vol. 53, pp. 217-235; September, 1949. 

(2) D. O'Dea, "Magnetic recording for the technician," Jour. SMPE, vol. 51, 
pp. 468-480; November, 1948. 

E. Masterson, "35-mm magnetic recording system," Jour. SMPE, vol. 51, 
pp. 481-488; November, 1948. 

G. L. Dimmick and S. W. Johnson, "Optimum high-frequency bias in mag- 
netic recording," Jour. SMPE, vol. 51, pp. 489-500; November, 1948. 

W. A. Mueller and G. R. Groves, "Magnetic recording in the motion pic- 
ture studio," Jour. SMPE, vol. 52, pp. 605-612; June, 1949. 

O. B. Gunby, "Portable magnetic-recording system," Jour. SMPE, vol. 52, 
pp. 613-618; June, 1949. 

S. W. Johnson, "Factors affecting spurious printing in magnetic tapes," 
Jour. SMPE., vol. 52, pp. 619-628; June, 1949. 

A New//1.5 Lens 

For Professional 16-Mm Projectors 



Summary To meet the growing demand for improved high-aperture 16- 
mm projection lenses, the Eastman Kodak Co. has announced a new series 
of f/1.5 lenses primarily intended for professional projectors. The lens is 
fundamentally of the Petzval type with field flattener. The resolving power, 
contrast and back focus clearance are substantially greater than hi present 
designs. The paper includes a historical outline of the development of 
lenses of this type, showing the progressive improvements hi aberration 
correction that have been attained in recent years. 

ONE HUNDRED AND NINE YEARS AGO, one year after Louis Jacques 
Mande Daguerre (1787-1851) had announced the first practical 
process of photography, the daguerreotype, another important con- 
tribution to the new science, was introduced. Joseph Petzval (1807- 
1891), a professor of higher mathematics at the University of Vienna 
who had undertaken the design of a new lens for photographic pur- 
poses, was ready to present the results of his computations, the fa- 
mous "Portrait Lens" (Fig. 1). The performance of the new lens, be- 
cause of its excellent correction of spherical aberration and coma in 
combination with the unusually high aperture of //3.4, was of such 
outstanding quality that it became the most prominent lens in the 
studios of all photographers. Besides the photographers, other 
groups, such as the manufacturers of optical instruments, as well as 
the contemporary opticians showed great interest in the new lens. 
The theoretical opticians of that time were impressed mainly by Petz- 
val's new methods of lens design. His theories concerning aberra- 
tions, astigmatism and curvature of field were indeed new, and his 

theorem, the well-known "Petzval Sum" ( = / - ) has been 

\R P t-Jn) 

* R p is the radius of field curvature near the axis of the optical system when it 
is free from astigmatism; and 

<f> is the surface power f J where n' and n are the indices of refraction and 

r the radius of curvature. 

PRESENTED: October 14, 1949, at the SMPE Convention in Hollywood. 





Fig. 1. Original Petzval Fig. 2. Dallmeyer (1866) Fig. 3. Voigtlander 
Portrait Lens (1840) Zincke-Sommer (1870) (1878) 

Fig. 4. Kodak Projection 
Lens //1.6 (1935) 

Fig. 5. Kodak Projection 
Lens f/lA (1939) 

Fig. 6. M. V. Rohr 
(Zeiss) (1911) 

Fig. 7. Kodak Projection 
Ektanon//1.6 (1938) 


Fig. 8. Kodak Projection 
Ektar//1.5 (1949) 




p /l.3 


-.5 +.5 -.5 | +.5 -75 I -K5 -.5 | +.5 -75 |+J5 -.5 I +.5 

(a) (I) (a) (A) (a) 

FIG. 7 FIG. 8 

Curves of spherical aberration, focal length differences and field curvatures for the 
lenses shown in Figs. 3 to 8 on the opposite page. 

340 W. E. SCHADE March 

an important "tool" to lens designers ever since. It has been said 
that with Petzval the era of serious lens design had begun. 

For twenty-five years the "Portrait Lens" was without any rival. 
Finally, in 1866 J. H. Dallmeyer 1 entered the competitive field with a 
modification of the Petzval lens, which was followed in 1870 by 
Zincke-Sommer with a similar design (Fig. 2). The modifications of 
the original Petzval lens consisted mainly in the reversed arrangement 
of the two single lenses composing the rear component. However, 
both designers were successful in increasing the relative apertures to 
//3 and//2.37 respectively. 

In 1878 Voigtlander 2 was granted a German and a British patent 
based on a further modification of the Dallmeyer and Zincke-Som- 
mer designs. He had succeeded in eliminating the airspace in the sec- 
ond component, and claimed that his lens, having only four air-glass 
surfaces against six in the other portrait lenses, would transmit more 
light and therefore render clearer images (Fig. 3). 

The spherical aberration (full line) and the focal length difference 
(dotted line) of this Voigtlander //3. 5 lens for a focal length of 100 mm 
is indicated in Fig. 3a. The under-corrected spherical aberration, 
however, could be easily reduced and would then be closely compa- 
rable to that of the original Petzval lens. 

The field curvatures for a focal length of 100 mm are shown in Fig. 
3b. The meridional field (dotted line) is flat, but the sagittal field 
(full line) is curved.* The longitudinal separation between the two 
fields, the astigmatism, increases rapidly with obliquity. 

With the approach of cinematography, it was this lens type that 
was selected to be used as a projection lens for 35-mm, 16-mm and 8- 
mm projection. Its merits of high contrast and excellent definition in 
and near the center, combined with simplicity and reasonable cost of 
manufacturing, have tempted many optical designers 3 to increase its 
usefulness with respect to higher relative apertures with improved 
aberrations and field curvature. 

Such a lens is the Kodak Projection lens shown in Fig. 4. It is of 
the Petzval-Voigtlander type with a relative aperture of //1. 6. The 
spherical aberration (Fig. 4a) is small and the meridional field is flat 
(Fig. 4b) . The performance of this lens is satisfactory. A lens of 2-in. 

* The meridional field is determined by oblique rays which enter the optical 
system in the plane of the drawing. 

The sagittal field is determined by oblique rays entering the optical system in a 
plane perpendicular to the plane of the drawing. 


focal length, which is the standard for 16-mm projectors, resolves 4 
about 100 lines/mm in the center, gradually falling off to about 20 
lines/mm in the corner of the 16-mm frame, at a semi-angle of Q^ . 

Another recent modification 5 is the Kodak Projection lens //1.4. 
To attain a relative aperture higher than //1. 6 the designer divided 
the second component into two, and was thus able to increase the 
relative aperture to as high as //1. 3. The spherical aberration, as 
shown in Fig. 5a, for an aperture of //1. 4 is small and the field char- 
acteristics are the same as those of the examples, Figs. 3 and 4. The 
lens was especially designed to be used in focal lengths longer than 2 
in. A lens of 3-in. focal length will resolve 90 lines/mm in the center 
and 40 lines/mm in the corner of the 16-mm frame. The projected 
image is of high contrast and also free from lateral color and distor- 


As early as 1874, the British astronomer Charles Piazzi Smyth 
(1819-1900) invented an ingenious device for flattening the curved 
field of the Petzval portrait lens. 6 He placed a negative lens of suf- 
ficient power in the focal plane of his portrait lens and achieved a great 
improvement in the field curvature and covering power. 

M. Von Rohr, the well-known historian, reports that the photo- 
graphs taken by Smyth with his Petzval-Smyth lens combination 
were of remarkable quality. For many years Smyth's simple but no- 
ble device seems to have been forgotten, or to say the least, neglected. 
In the year 1911, Von Rohr 7 (Carl Zeiss) introduced a lens combina- 
tion comprising a portrait lens to which a negative field-flattener lens 
was attached near the focal plane (Fig. 6). 

His design differs from the original Smyth arrangement in several 
ways. In order to obtain a useful back-focus clearance, he had to 
design the entire system as a unit, because when the field flattener is 
moved away from the exact focal plane, it immediately begins to af- 
fect all the aberrations of the main lens. To obtain another degree of 
freedom in his work of design, he introduced an airspace between the 
two elements of the front component of the portrait lens and was able 
to achieve a relative aperture as high as //1. 8. The lens consists of 
five single elements, separated by air. The spherical aberration (Fig. 
6a) is well corrected, but the zone (//2.6) shows a considerable amount 
of residual under-correction. Instead of flattening at least the merid- 
ional field, the designer seemed to be more interested in freedom from 

342 W. E. SCHADE March 

astigmatism with the result that both fields were curved. Neverthe- 
less, the improvement over the original Petzval portrait lens is im- 

The Smyth principle has been applied in the well-known Kodak 
Projection Ektanon lens, 8 shown in Fig. 7. It consists of the modified 
Petzval type and a field flattener, with a relative aperture as high as 
//1. 6. The aberrations (Fig. 7a) are such that this lens of 2-in. focal 
length resolves 90 lines/mm in the center of the projected image. 
Astigmatism and the curvatures of the meridional and sagittal field 
(Fig. 7b) show a considerable improvement over the lenses without 
the field flattener, and consequently there is a more satisfactory per- 
formance. A lens of 2-in. focal length will resolve 50 lines / mm in the 
corner of the 16-mm frame. 


The progressive improvement of lenses of the Petzval type, espe- 
cially those embodying a field flattener, has been described in the pre- 
vious paragraphs. 

However, the demand for a 16-mm projection lens, comparable with 
the best used in 35-mm projection, is steadily increasing. Next to a 
high relative aperture it is imperative that the resolving power of such 
a lens should be of a higher order than ever before. These require- 
ments have led to a new design, the Kodak Projection Ektar //1. 5 
(Fig. 8). 

This lens is fundamentally of the modified Petzval type to which a 
field flattener has been attached. To surpass the performance of pre- 
vious projection lenses, the designer has not only added one element to 
the front component to reduce the zonal spherical aberration, but has 
also taken advantage of the new high-index glasses 9 of which elements 
3 and 4 consist. 

The zonal spherical aberration (Fig. 8a) has been reduced to a mini- 
mum. Astigmatism and field curvatures (Fig. 8b) also show a distinct 
improvement which has to be credited to the use of high-index low-dis- 
persion glasses. These have already proved their value in other opti- 
cal systems. 10 In the present case the use of these glasses relieves the 
field flattener of some of the burden of Petzval sum correction. In 
this way the flattening of the sagittal field (formerly the worse of- 
fender) has been solved satisfactorily. The relation of the two fields 
to one another is such that, by a slight bending of the field flattener, 
astigmatism could be brought to zero with both fields flat. However, 




since the entire oblique pencil had to be considered and the oblique 
spherical aberration had shown a tendency to over-correction, a slight 
inward displacement of the sagittal and meridional focal points was 

Fig. 9. A series of Projection Ektars. 

The performance of this lens is indicated in the curves in Fig. 8a and 
8b. It resolves better than 90 lines/mm over the entire 16-mm frame 
with a perfectly flat field. With respect to contrast, the lens has 

344 W. E. SCHADE 

proved to be a worthy descendant of its famous ancestor, the Petzval 

It was decided to manufacture a series of these //1. 5 Projection 
Ektar lenses in 2.062-in. barrels, the focal lengths being in geometrical 
progression as originally recommended by Rayton. 11 The choice of 
common ratio was based on the decision to allow two intermediate 
lenses to fall between the standard 2-in. and 3-in. sizes. The new focal 
lengths thus became 2J4 and 2% in., at a common ratio of about 12% 
(Fig. 9). If this series were continued one step further in the long di- 
rection it would fall at 3^ in., but it was felt that this is probably too 
long for most civilian applications. However, longer sizes can, of 
course, be made if required. 

In every case, the resolving power of 90 lines/mm has been rigidly 
maintained in manufacture, over the entire area of a flat 16-mm pro- 
jector gate. 


(1) J. H. Dallmeyer, U.S. Pat. 65,729 (1867). 

(2) F. von Voigtlander, Ger. Pat. 5761 (1878) or equivalent Br. Pat. 4756 

(3) W. H. Repp, U.S. Pat. 1,479,251 (1922); A. Warmisham, U.S. Pat. 
1,484,853 (1922); C. Graf, U.S. Pat. 1,610,514 (1925); R. Richter, U.S. Pat. 
1,843,519 (1931); and A. Warmisham, Br. Pat. 485,096 (1936). 

(4) R. Kingslake, "Resolution tests on 16 mm projection lenses," Jour. SMPE, 
vol. 37, pp. 70-75; July, 1941. 

(5) W. Schade, U.S. Pat. 2,158,202 (1939). 

(6) W. Merte", R. Richter, and M. von Rohr, Das Photographische Objectiv, pp. 
10-11; Springer, Wien, 1932. 

(7) W. Merte", R. Richter, M. von Rohr, Dos Photographische Objectiv, pp. 326, 
329; Springer, Wien, 1932. For other lenses embodying a field flattener, see: 
H. D. Taylor, Br. Pat. 127,058 (1917); A. Warmisham, Br. Pat. 492,311 (1937); 
and F. W. Gehrke, U.S. Pat. 2,187,780 (1938). 

(8) D. Wood, U.S. Pat. 2,076,190 (1934). 

(9) R. Kingslake and P. F. De Paolis, "New optical glasses," Scientific 
Monthly, vol. 68, p. 420; June, 1949. 

(10) G. H. Aklin, "The effect of high index glasses on the field characteristics of 
photographic objectives," Jour. SMPE, vol. 38, pp. 841-844; October, 1948. 

(11) W. B. Rayton, "A proposed new series of standard focal lengths for motion 
picture projection objectives," Jour. SMPE, vol. 15, pp. 270-280; July, 1930. 

The Metal-Diazonium System 
For Photographic Reproductions 


Summary The metal-diazonium system for photographic reproduction, 
which has been developed in Philips Laboratory in Eindhoven in the course 
of the last few years, is based on the discovery that when a solution of a di- 
azonium compound and a metal salt, say mercurous nitrate, is exposed to 
light, atomic metal in casu mercury is separated. The "latent" 
mercury image thus obtained can be transformed by physical development 
into a silver image and intensified. The light-sensitive system is obtained 
in the form of a film or sheet by impregnating a suitable carrier, say a strip 
of cellophane 40 /z thick, in a homogeneous solution of the said materials. 
The metal-diazonium system possesses an extremely high resolving power 
(> 1000 lines/mm) and allows of working with a very high gamma (6-8) 
while on the other hand low gammas (1-2) can also easily be obtained by 
varying external factors, viz. the moisture content or the intensity of exposure. 
The light-sensitivity of the system is in cellophane several times (in paper 
some tens of times) greater than that of the usual diazotype printing papers. 
This system, which was originally intended only for producing distortion- 
free copies of Philips-Miller sound film, lends itself excellently, inter alia, 
for the copying of picture-sound films, thanks to the external variability of 
the gamma. The impregnating of the cellophane base takes place on print- 
ing machines designed for the purpose, while at the same time these ma- 
chines are fitted with a device for regulating the moisture content of the base. 
The good photographic properties of the system and the very low cost of ma- 
terials open great prospects for its application on a large scale in all sorts of 
fields. In addition to the sound film and the picture-sound film also the 
field of micro and macro documentation is regarded as an important domain 
for the application of the system. 

rriHE metal-diazonium system is a new light-sensitive system that 
JL has been worked out by a group of scientists in the Philips Lab- 
oratory at Eindhoven. It possesses a number of remarkable proper- 
ties making it exceptionally suitable for the photographic reproduc- 
tion of pictures as well as of sound. Something has already been 
said about this system in a previous article, 1 where it was shown by 
a comparison with the usual methods of reproduction what place the 
new system could occupy. 

Here we shall give a more detailed description of the fundamental 

1 C. J. Dippel and K. J. Keuning, "Problems in photographic reproduction, 
in particular of sound-films," Philips Tech. Rev., vol. 9, no. 3, pp. 65-72; 1947. 
REPRINTED FROM: Philips Technical Review, vol. 9, no. 10, pp. 289-300. 



principles of the metal-diazonium system and after going more deeply 
into its properties show how it is realized and practiced, then con- 
cluding by dealing with a number of perspectives for the application 
of this new reproduction material. 

The Light-Sensitive Material 
Diazonium salts have the general chemical formula 

[R - N=N] + X- 

in which R is an aromatic radical and X some anion. These com- 
pounds have been used for quite a time already for photographic re- 
production methods. They are most familiar and most widely used 
in diazotype, a light-printing process for the multiplication of tech- 
nical drawings and suchlike (tracings) made on transparent material. 
Just as with all reproduction methods based on diazonium com- 
pounds, the fundamentals of this light-printing process lie in the fol- 
lowing properties : 

(a) Coupled with certain phenols or amines, the diazonium salts 
form what are known as azodyestuffs. 

(b) When a diazonium salt is exposed to light in the presence of 
water (water vapor) dissociation takes place and nitrogen is released. 
Schematically the reaction takes place as follows: 

RN 2 X + H 2 O -f h v -* l.d.p. + N 2 + HX. (1) 

The light-decomposition product (l.d.p.) thus formed is no longer 
capable of forming a dyestuff. If, therefore, one exposes a "diazotype 
paper" on which a tracing is laid and then causes the reaction, (a) 
above, to take place, a dyestuff is formed only on those parts of the 
paper that have not been exposed to the light. In this way a positive 
copy is obtained direct from the tracing. 

The difficulties and limitations referred to in the previous article 1 
as being inherent in photographic reproduction with the usual silver- 
bromide and silver-chloride systems exist in a still higher degree with 
the diazonium processes hitherto known. They have not met with 
any appreciable success, for instance, in the reproduction of picture- 
sound films. 

The new photo-chemical system described here is based on the dis- 
covery that the light-decomposition products obtained from certain 
diazonium compounds according to equation 1 are capable of releas- 


ing the metal from suitable metal salts, for instance mercury from 
mercurous nitrate, gold from aurochloride, etc. As diazonium com- 
pound one may use for instance o-cresol-diazonium-sulphonic acid : 

OH + 

so 3 - 

When an aqueous solution of this compound and, for example, mer- 
curous nitrate, Hg 2 (N0 3 )2, applied in a thin flat layer is exposed with 
light of a short wavelength (e.g. 3650 A) one may observe under the 
microscope small drops of metallic mercury at the places struck by the 
light. Thus a faint "mercury picture" is formed which might be com- 
pared to the "latent" picture in the usual method of photography with 
silver halogenides. And here, too, the "latent" picture can be in- 
tensified and made durable by developing. Contrary to the custom 
with silver halogenide photography, however, a so-called physical de- 
veloping process is applied with our system. In view of the important 
part this developing method plays in the metal-diazonium system a 
separate section will be devoted to it below. 

In order to show properly the difference from the old diazonium 
processes the point is stressed that only the second of the two char- 
acteristic properties of diazonium salts mentioned above, viz. the light 
decomposition, is utilized in the metal-diazonium system, the other 
property formation of dyestuffs playing no useful part in our 
system. The diazonium compound left on the unexposed parts is 
removed, while the exposed parts turn black (separation of metal). 
Thus we get instead of a positive copy a negative copy, as is the case 
with silver halogenides. 

As to the practical realization of the system we shall revert to this 
later, but it is necessary to say something here about the manner in 
which the thin layer of the light-sensitive material is obtained. In 
this respect the new system differs fundamentally from the silver 
halogenide systems. With the latter systems an "emulsion" (more 
correctly: a suspension) of the light-sensitive substance, e.g. crystal- 
line silver bromide, is made in gelatine and after a complicated and 
most precise ripening process this emulsion is cast on a celluloid 
film or a glass plate. In our case, on the other hand, a homogeneous 
solution is made of the diazonium compound and the metal salt and 
a suitable carrier is saturated with it. As carrier one may use for 


instance paper or, as we have done in the most important applications, 
a transparent base of regenerated cellulose. This latter material 
(which is more commonly known under the name of cellophane and 
is widely used as packing material for shop goods) is used by us in the 
form of a reel of film 0.04 mm (0.0016 in.) thick. But in principle 
one may also use with our system a gelatinous layer on celluloid or 
glass. One may also use the cellulose acetate film (so-called safety 
film) commonly applied for substandard films and by saponification 
make the surface suitable to absorb the aqueous solution of our light- 
sensitive system. 

Physical Developing Process 

What takes place in the so-called physical developing process in 
photography may be resolved into a phenomenon often met with in 
nature and in technology and which might be described as follows. 
When particles from an over-saturated solution or vapor begin to 
precipitate they show a preference for places where certain "nuclei" 
are present. If these nuclei are distributed locally in such a way as 
to form a "picture," maybe so faint that the eye cannot see it, then 
the picture is intensified by this selective precipitation of particles and 
may thereby be made visible. 

A familiar phenomenon known to everyone is the picture formed 
by condensation on a smudged glass window when the window is 
breathed upon or cold air blows upon it from one side, water vapor 
condensing on the traces of dirt acting as nuclei and making visible 
figures. 2 Another example that may be given is that of the Wilson 
camera, where droplets of water from an oversaturated vapor con- 
dense on electrically charged particles and thus make visible the path 
followed by an ionizing particle. 

Turning particularly to photography, we see that both chemical 
and physical developing methods are in use, the latter more par- 
ticularly in cases where negatives have to be intensified. Let us first 
consider what ordinary chemical developing comprises. A silver 
bromide film or plate is placed in a developing bath containing a re- 
ducing agent, for instance metol or hydrochinon, in a generally alkaline 
solution. The grains of silver bromide, in which silver nuclei have 

2 In the familiar game of writing or drawing on a glass window with a wet 
finger, when some days later the writing is made visible by breathing upon it, 
we have examples of negative pictures; the window is more or less uniformly 
covered with "dirt," which is removed by the finger (or at least partly so) or else 
a thin film of grease is left behind, so that less vapor is condensed on the writing 
than on the rest of the window. 


been formed by the exposure (groups of say 100 atoms of metallic 
silver), are reduced by the developer entirely to silver, while the un- 
exposed or too weakly exposed grains remain untouched. After this 
developing, as is known, the remaining silver bromide is removed 
with the aid of sodium thiosulfate (fixing). 

If the picture thus obtained is not dense enough, it can be intensi- 
fied by physical development. The film is placed in a weak acid 
solution containing silver nitrate in addition to a reducing agent, for 
which metol or hydrochinon can again be used. In the solution the 
silver nitrate is gradually reduced to silver, the solution becoming 
oversaturated, as it were, with atomic silver, which precipitates pref- 
erably on those places where "silver nuclei" are already present, i.e. 
on the exposed parts of the film. 3 

It is perhaps of interest to note that this physical developing process can be 
applied also to the "latent," not yet chemically developed, image on the exposed 
silver bromide film. One can start by fixing the exposed film, thus removing all 
the silver bromide and leaving only the silver nuclei formed by the exposure, the 
latter then acting as nuclei for the subsequent physical development as described 

The oldest photographic process, daguerreotype, was likewise based upon a 
physical method of developing; an image was formed on an iodized silver plate 
and the exposed plate was treated with oversaturated mercury vapor. Mercury 
was thereby condensed on the exposed parts of the plate and the picture became 

After these examples the method of physical developing applied 
with the metal-diazonium system does not need much more explana- 
tion. The carrier, containing for instance o-cresol-diazonium-sul- 
fonic acid and mercurous nitrate, is placed after exposure in a suit- 
able solution of silver nitrate and a reducing agent. Reaction then 
takes place between the locally formed metallic mercury and the silver 
nitrate, the mercury dissolving and silver precipitating at those 
places, the latent mercury picture thus being transformed into a 
latent silver picture. Moreover, by reduction metallic silver is grad- 
ually formed in the solution. The local metal deposits already pres- 
ent act as nuclei upon which more and more silver is deposited as new 
silver is formed in the solution, thereby developing the picture. 

3 In this case the physical developing process comprises also a chemical proc- 
ess. From this it appears that the name "physical development," which we use 
because it has already been introduced, does not express the essential difference 
from the actual chemical process of developing. The difference as we see it lies 
in the fact that in chemical development the metal from which the picture is 
built up is already present in the appointed place prior to developing in any form 
whatever, whereas in the physical development it is only brought into its ap- 
pointed place by the developing process. 


One of the essential factors in this process is the rapidity with which 
the silver is formed in the solution by reduction. If this takes place 
very rapidly one gets instead of a selective deposit a more or less 
evenly spread deposit of silver, a phenomenon that can even be 
turned to advantage for making homogeneous silver mirrors. By 
giving the solution, for instance, a suitable degree of acidity one can 
regulate the speed of the spontaneous reduction of the silver nitrate 
and cause practically all the silver formed during the developing proc- 
ess to precipitate on the nuclei of the picture, to the exclusion of al- 
most all undesired precipitation of silver on the unexposed parts. 
The negative may not be kept in the developing bath longer than the 
time taken for developing, because the developer is an unstable system 
and liable in time to cause a spontaneous flocculation of all the silver, 
which then precipitates anywhere. 

We will not conclude this explanation of the principles of the metal- 
diazonium system without remarking that here only a very rough 
and greatly simplified representation of the processes has been given. 
To understand the connection between the photographic properties 
obtained and the numerous variable factors of the system it has been 
necessary to study deeply the mechanism of the light decomposing 
reaction (equation 1) and of the developing process. It may be pos- 
sible at a later date to go more deeply into the problems arising, some 
of which are still unsolved. 

Resolving Power 

In the article previously quoted 1 it was explained that for photo- 
graphic reproduction, especially of picture-sound films, a high re- 
solving power is favorable. However, there is a limitation in this 
respect, in that it tends to spoil the quality of reproduction, (unsharp 
pictures, distortion of high frequencies) ; or, when trying to avoid this 
drawback, it necessitates a greater length of film (the picture may not 
be too small), the same applying for the speed of the film on account 
of the sound. 

The positive films commonly used for copying picture-sound films 
have a resolving power of between 50 and 75 lines/mm, so that when 
projecting gratings finer than what corresponds to this number of 
lines per mm the lines run into each other and therefore can no long- 
er be made visible separately. Some new Kodak films for special 


purposes have a resolving power as high as 160 lines/mm. The 
metal-diazonium system is quite capable of resolving 1000 lines /mm. It 
is even possible that its resolving power is still greater, for the opti- 
cal means used in determining this power have themselves a limited 
resolving power, which in our case did not reach further than the said 
limit of 1000 lines/mm. 

In other laboratories too, i.a. Kodak, it has for some time been possible to make 
reproductions with an extremely high resolving power, of the order of 1000 lines/ 
mm. 4 Collodion plates or so-called Lippmann emulsions are used. Except for 
the already known application of wet collodion plates for making autotypes, as 
far as we know none of these processes has been suitably developed for general use. 

The high resolving power of the metal-diazonium system is due for 
a large part to the fact already mentioned that the light-sensitive 
material is not used in the form of an emulsion but in that of a homo- 
geneous solution; the unexposed system is quite free of grains, so that 
there is very little diffusion of light in the sensitized material. If a 
large quantity of light reaches one point then in the circle round about 
that point where undesired light reaches, owing to diffusion, the 
quantity of that light is extremely small. Moreover, and this is the 
second cause of the high resolving power, that circle (the diffusion 
halo) remains extremely small owing to the fact that with the chosen 
concentration of the chemical components in the light-sensitive 
solution the active light is very strongly absorbed. Consequently the 
diffused light does not reach far. 

Owing to the strong absorption the direct light thrown upon the 
carrier upon exposure stays in the top layer. Consequently the metal 
forming the picture is limited to a thin layer. This is likewise of im- 
portance for the high resolving power, for it is not sufficient that de- 
tails are recorded well separated in the carrier they must also be 
reproduced separately either for copying or for projection. Owing to 
the fact that one never uses perfectly parallel light beams, the thicker 
the picture layer the more details are lost. 5 

4 See, for instance, J. Sd. Instr., vol. 18, pp. 66-67; 1941, where an account 
is given of the "Kodak maximum resolution plate," which can resolve 600 or even 
1200 lines /mm. Similar results have been reported by H. Frieser, Z. wiss. Phot., 
vol. 40, p. 132; 1941. For older methods, see E. v. Angerer, Wissenschaftliche 
Photographic, p. 136 et seq. Akad. Verl. Leipzig, 1931. 

5 Also hi the old diazotype processes based on diazonium compounds the ma- 
terial is free of grains, but the resolving power is not very high, because, i.a., the 
picture is rather thick: owing to the relatively small absorption a thick layer of 
dyestuff is necessary for adequate "density." Furthermore the formation of the 
dyestuff is a relatively slow reaction, so that after the exposure a noticeable diffu- 
sion takes place before the (fixed) molecules of the dyestuff are formed. 




Figure 1, the magnified reproduction of a micro-document made 
with the new system, demonstrates the high resolving power. 

The Gamma Value 

Every photographic picture, be it positive or negative, has a cer- 
tain characteristic density curve which indicates the density obtained 
for any exposure E = I t (I luminous intensity, t exposure time) . The 
density D 6 is usually plotted as a function of log E; see for instance 
Fig. 2. For many light-sensitive materials the density curve shows a 

The 'la ft 

r '. ha vc* 
r* *d| tirf t c| 

-nr Miller 

mll far 



reduces I 

Fig. 1. A reproduction, linearly enlarged 
about 250 X, of a piece of a micro-document 
recorded with the metal-diazpnium system in 
cellophane. The size of this piece in the micro- 
document was 0.012 X 0.017 cm and the 
height of the letters 12 microns. The good defi- 
nition of the reproduced letters gives an idea 
of the exceptionally high resolving power 
of the system ( > 1000 lines /mm). 

log ft- 1) 

Fig. 2. Example of a density curve D = 
/(log /<) of a photographic picture. The 
maximum slope of the curve lying in the 
practically rectilinear part in the middle 
is the gamma. 

more or less extended, practically rectilinear part, where at the same 
time the slope of the curve is greatest. The whole curve is in fact 
characterised by this slope, the gamma, because this is decisive for 
the gradation in the reproduced picture (reproduction of the shades 
of brightness) . It is well known that w r ith the usual silver halogenide 
systems the value of the gamma depends largely upon the emulsion 
and is further particularly determined by the conditions of develop- 
ing (temperature and composition of the developing bath, time 
taken in developing) . Common values in practice vary from 0.5-2.5. 
When applied in a suitable manner, as will be defined below, the 
6 Defined as D = log io/i, where i is the portion of an incident quantity of light 
io that the blackened plate or film allows to pass through. 


metal-diazonium system has very much higher gammas, e.g. 6-8. 
In the previous article 1 it has been explained what important 
advantages this offers for instance for sound reproduction. Since 
with a high gamma a relatively small reduction in the exposure in- 
tensity is sufficient to bring about a transition from the "greatest 
density" to the "smallest density," only the innermost part of the 
circle diffusion halo already referred to is noticeably blackened. 
Thus the high gamma promotes a high resolving power and therefore 
in the case of sound reproduction promotes good reproduction of the 
high frequencies. The high gamma yields particular advantages in 
the copying of Philips-Miller film, as we have seen in the previous 
article; the "lens effect" arising in this process is rendered harmless 
without any other measures being necessary. 7 

External Variability of the Gamma 

In sound reproduction by the amplitude system, advantage can 
safely be taken of a high gamma, because in principle we have only to 
do with two densities, a very low one in the transparent sound track 
and a very high one for the rest of the film. There are all sorts of other 
applications where the same holds, e.g. in micro and macro documen- 
tation, to be discussed below. In picture reproduction, on the other 
hand, a whole series of shades of brightness (half tones) has to be 
reproduced by corresponding, continuously varying densities. To 
attain this it is necessary to satisfy the Goldberg condition, 8 as a re- 
sult of which in picture reproduction one has to work with relatively 
low gammas, e.g. 1.5-2.5. 

Now such gammas and still lower values can be obtained quite 
easily with the metal-diazonium system. This is possible not only by 
a suitable choice of the composition of the system but also by making 
use of the following important feature: the gamma of a metal-diazo- 
nium-cellophane system can be greatly influenced by the moisture content 
of the film during exposure and likewise by the duration of the exposure. 
The very high gammas referred to under the previous heading occur 

7 Also the so-called wedge effect with the Philips-Miller film is rendered harmless 
by the high gamma. When the transparent sound track in a "Philimil" film is 
cut with the wedge-shaped chisel, which is a characteristic feature of the Philips- 
Miller system, an oblique edge is left on the top layer, and when copying this 
results in a gradual transition from the maximum to the minimum density. In 
itself this does no harm, but aberrations arise if there is any variation in the thick- 
ness or the density of the covering layer. The higher the gamma, the narrower 
this edge transition shows on the copy and the smaller the aberration. 

8 See the explanation in the article quoted in footnote 1. 




when the cellophane film is dry, that is to say when its moisture con- 
tent is not more than say 15% by weight, and when the film is exposed 
with a great intensity (and corresponding short exposure time). If 
the moisture content is increased to 25-30% by weight, then one gets 
the low gammas required for good picture reproduction. 

Figure 3 indicates the relation between the gamma value and the 
moisture content of a cellophane film for certain concentrations of the 
ingredients and under certain conditions of exposure, developing, 
etc. There the following refinement has been applied. Instead of 
showing the gamma indicating the maximum slope of the density 
curve, the values gi, g%, g s have been plotted of the average slope that 
the density curve assumes in three consecutive density intervals of 

Fig. 3. Effect of the moisture 
content of the cellophane film 
upon the gamma of the mer- 
cury-diazonium system in the 
film for a certain concentration 
of the ingredients and a certain 
manner of exposure and develop- 
ing. The lines plotted repre- 
sent the average slopes gi, g 2 , g* 
of the density curve in three dif- 
ferent density areas; g 3 is prac- 
tically equal to the maximum 
slope 7. 

the most importance in practice, viz. between D = 0.05 and 0.5 (high 
lights); D = 0.5 and 1.0 (intermediate tones) ; D = 1.0 and 1.5" (shad- 
ows). The value g^ is generally practically equal to the maximum 
slope defined as 7. 

This defined description of the gradation, often applied with silver halogenide 
systems and sometimes even extended to five intervals, is desired when one has a 
group of density curves which do not all have a rectilinear part with maximum 
slope in the density area required. This is also frequently the case with the 
metal-diazonium-cellophane system, with which it is possible to get a large variety 
of density curves under the influence of the numerous variables. 

The fact that the gamma, or in other words the density curve, de- 
pends upon the duration of exposure t means that the density D is 
no longer a function of the product I-t but of / and t separately. 
Therefore instead of a two-dimensional density curve, to describe 
fully the photographic behavior of the system we need a solid figure 
in which the curved plane D = /(/, t) is represented. Such a plane 
is drawn in perspective in Fig. 4. It must be borne in mind that by 


varying the moisture content, the conditions of developing, etc., a 
different plane is obtained every time. The "density curve" of a 
picture taken with an exposure time t\ is the cross section of the 
density plane parallel to the D and / axes which is intersected by the 
t axis at h. From Fig. 4 it is clearly seen that for different values 
of t one obtains density curves with a different slope (gamma). A 
number of these curves are drawn in Fig. 5, while in Fig. 6 the slopes 
Qi> 02, 03 ( T) of these curves are plotted as a function of t. 

From this it also follows that the relation between log / and log t for a constant 
density D (horizontal cross sections of the curved plane at different heights; 
see Fig. 4) is given by lines the slope of which must gradually change with varying 
D. Whereas in the simplest case, which was assumed in Fig. 2, D was only a 
function of I-t so that for the given D we had log I + p log t = constant with p = 1, 
in our case as a rule p ^ 1. This is the familiar Schwarzschild effect. Moreover, 
according to the foregoing, in our case the Schwarzschild exponent p (slope of the 
log /-log t lines in the area where these may be regarded as straight) depends also 
upon the density. 

Figures 3 and 6 give a clear picture of the great variability of the 
gamma in the metal-diazonium system. Now it is important to 
note that this is an external variability: on one and the same material 
and with one developing process we can make copies with a high 
gamma and with a low gamma. This creates not only the possibility 
already mentioned of copying both sound and pictures each with the 
most favorable gamma, but also an entirely new possibility of copy- 
ing sound and picture side by side on 'one film and developing them 
together without necessitating a compromise in the gamma such as 
is characteristic for the present picture-sound film technique (see 
the previous article 1 ). All that is necessary is either to select the 
exposure intensities separately for the copying of the pictures and 
for the copying of the sound track, or else to vary the moisture con- 
tent of the cellophane band between the two places where in the 
printing machine first the picture is copied and farther on the sound 
track is copied (it may be that both measures have to be applied 
together) . 

In a similar manner we can also print on one film the pictures of 
different scenes with a different gamma while requiring only one de- 
veloping process! This offers the possibility of correcting any ex- 
posure or developing variations in the negatives. 

Light Sensitivity 

The light sensitivity of the metal-diazonium system depends not 
only upon the choice of the ingredients, etc., but also varies with dif- 




log I 

Fig. 4. The gamma of the metal-diazonium system also depends, i.a., upon the 
intensity of exposure /. Therefore the density D is not fully determined by the 
product 1-t; the photographic behavior of the system has to be described by a 
solid density plane D = /(/, t}. This plane is drawn here for given concentra- 
tions, moisture content, developing, etc., in perspective. Log 7 and log i are 
plotted as independent variables. 

Fig. 5. From the density plane 
drawn in Fig. 4 it is possible to 
find for any exposure time t the 
density curve D = /(/) of the 
picture obtained with that ex- 
posure time, by taking a cross 
section of the plane perpendicu- 
lar to the i-axis at the level of 
that time. Such cross sections 
are represented here for various 
exposure times. It is seen that 
curves are obtained with dif- 
ferent slopes (different gammas). 


Fig. 6. The slopes g it g z , g 3 
of the curves of Fig. 5 plotted 
as functions of the exposure 
time t . 


ferent carriers. When cellophane is used the sensitivity may be sev- 
eral times greater than that of the known diazotype papers. Yet with 
paper as a carrier the sensitivity of the system is greater by a factor 
of 10. For instance with mercury as metal the metal-diazonium 
paper can easily be made 20-25 times as sensitive as the positive 
diazotype papers, a gain which just makes it possible to produce 
enlargements by the light-printing process (see the last section). 
Still this is a factor 10 4 below the sensitivity of silver bromide en- 
largement papers. 

There is, therefore, no question that the metal-diazonium system as 
at present developed could compete with the materials commonly used 
for photographic recording. As a matter of fact the spectral zone in 
which the metal-diazonium system is sensitive is too limited for this 
purpose. The sensitivity of the mercury-diazonium system lies 
mainly in the near ultraviolet with a maximum in the vicinity of 
3900 A and not extending beyond the bluish green (about 5000 A). 

For application as reproduction material, however, in most cases 
neither the low sensitivity nor the limitation of the spectral area con- 
stitutes any objection. One only needs to use for the copying process 
a light source possessing great luminous intensity in the said range of 
the ultraviolet. Super-high-pressure mercury lamps with water 
cooling are excellently suited for this purpose. With these lamps 
the intensity of light that can be reached when exposing the film is 
so great that a film can be copied at fairly great speeds, for instance 
20 meters per minute. 

An important advantage of the limited spectral sensitivity lies in 
the fact that the whole working of the system can take place under a 
bright sodium light, since in the wavelength range of the sodium 
lines (5890 A) the sensitivity is practically nil. 

A great deal of work has been done in this laboratory in respect to the question 
as to what determines the sensitivity of the metal-diazonium system. In the par- 
ticular case where mercury is employed as a metal it was demonstrated that the 
process of the formation of the latent mercury picture has a quanta yield of about 
50%; for an average of two exposure quanta one atom of metallic mercury is 
formed. This means that the "primary" sensitivity is of the same order as that 
of the silver halogenide systems. According to the conclusions provisionally 
reached from an investigation into the highly complicated mechanism, the cause 
of the so much smaller resulting sensitiveness of our system is to be sought rather 
in the further history of the nuclei to be developed; the metallic atoms combine to 
form larger particles. As already stated, under a very strong magnification 
(about 800 X) these particles may be seen in the latent picture. Thus in this 
stage the metal is rather coarsely dispersed and in the physical developing process 
the resultant density is all the less according as the (given) quantitj 7 " of metal of 
the latent picture is more coarsely distributed. 




The influence of the moisture content and the intensity of the exposure upon 
the gamma is also closely related to the history of the primary metallic atoms and 
of the metallic nuclei prior to developing. 


When talking of durability in the case of the metal-diazonium sys- 
tem we have to differentiate between the exposed and the unexposed 
state. The unexposed system appears to have as yet too little du- 
rability for a sensitized cellophane film, for instance, to be kept in 
stock until one has need of it. For the most important applications 
that we have in mind, however, this forms no serious objection, as 
will be made clear below. 

Fig. 7. Microtome cross section of a sensitized cellophane film 40 microns 
thick after exposure and developing. The particles of silver from which the pic- 
ture is built up lie in a thin layer a few microns below the surface. 

After exposure and developing one has a picture that will keep prac- 
tically indefinitely : as already mentioned above, the definitive picture 
consists of metallic silver, just as is the case with silver halogenide 
systems, and thus is perfectly proof against atmospheric influences 
and light (such contrary to the dyestuff pictures of the old diazotype 
papers). Furthermore, the picture on a cellophane film is protected 
in a peculiar manner against mechanical damage: when a micro- 
tome section of such a film is cut and examined under the microscope 
it will be seen that the extremely thin layer containing the metallic 
silver of the image (see above) does not lie on the surface of the film 
but a few microns below it; see Fig. 7. We cannot go into the ex- 


planation of this here, but it has the welcome practical advantage that 
the actual picture is protected against scratching, etc., by the thin 
layer of clear cellophane covering it. 


It is to be expected that the cost of the metal-diazonium system 
will be relatively low, a factor that will prove to be of great weight for 
various applications. How is it that the system turns out to be so 
economical? In the first place there is the choice of the carrier, cello- 
phane being the cheapest material imaginable for this purpose. An- 
other important factor is the limited consumption of silver, it being a 
characteristic of the physical developing method that the sensitive 
material itself need not contain any silver; the silver comes from the 
developer and is added to the negative, and the developing can be so 
controlled that not much more silver need be used than is necessary 
for building up the ultimate silver picture. With the usual silver 
halogenide systems, on the other hand, a very high percentage of 
silver is wasted; it comes out of the exposed film in the fixing bath 
and cannot be recovered except at considerable expense. 

Of further importance is the very simple method of manufacture, 
as will be explained under the next heading. 

Yet another advantage to be mentioned in connection with the 
economy of the metal-diazonium cellophane system is the saving in 
volume and the attendant ease of storage and transport. A cello- 
phane film 40 microns thick and say 300 m long winds up into a reel 
about 13 cm in diameter, whereas a normal celluloid film of 300 m 
length forms a reel 26 cm in diameter. 


The metal-diazonium system can be realized in quite different ways 
according to the use intended and the carrier employed. As a typical 
example we will consider here the application of the system for the 
copying of picture-sound films with cellophane as the carrier. 

Just as is the case with the common silver bromide films, so with the 
cellophane film the actual copying process takes place on a machine 
where both the negative and the positive films are caused to pass along 
under a lamp simultaneously. Now we have already said that the 
cellophane film is sensitized by impregnating it in its entirety with 
the solution containing the light-sensitive system. Further it has 
been stated that the cellophane film sensitized in this manner has only 





Fig. 8. Printing machine on which a cellophane film is first impregnated in 
the light-sensitive solution, then dried to the desired moisture content and after 
that exposed. The film is fed in from the right, impregnated in the bath at the 
top on the right, dried in the vertical tube and printed on the drum in the middle 
at the bottom of the photo, where it is brought into contact with the original film 
to be copied, the two films passing underneath the lamp simultaneously. (For 
drying the film when running at a high speed several tubes were used; with the 
latest machine drying is done by high-frequency heating.) 


a limited durability. The difficulties that this might involve have 
now been overcome in a very simple way by combining the sensitizing 
process with the printing process, the impregnating of the cellophane 
film and the exposure taking place on the same machine in succession. 

The fact that we have here an extremely simple and economical 
modus operandi is quite evident when comparing it with the manu- 
facture of silver bromide film, where the preparation of the carrier, 
the preparation and ripening of the emulsion, the casting of the emul- 
sion on the celluloid film and later the copying are all done in separate 

The first mentioned method has also been found to improve con- 
siderably the reproducibility of the properties of the film. 

Between the impregnating and the exposing of the cellophane film 
this has to be dried to a moisture content corresponding to the desired 
gamma. Further, it must be possible to reduce still further the mois- 
ture content between the exposure of the picture and that of the sound 
track on the film if such should be desired. This can be done by 
passing the film through a tube with conditioned air, as seen in Fig. 8, 
or by means of high-frequency heating. 


The system is still too young to allow of any data being given as to 
its applications, but the experience so far gained with it opens such 
interesting perspectives that a brief outline of some of its possible uses 
may well be given here. 

Sound Film 

A stereophonic sound track has been made on 7-mm cellophane 
films by copying a stereophonic Philips-Miller film. 9 Thanks to the 
sharp definition of the mechanically recorded original and the high 
resolving power of the copying material an exceptionally good quality 
of sound is obtained, in respect to the reproduction of the high fre- 
quencies and the absence of nonlinear distortion (see the article quoted 
in footnote 1 ). This good quality of reproduction together with the 
enhanced "naturalness" obtained by stereophony seem to us to con- 
stitute the requisites for imparting to "mechanical" music the original 
musical character. 

9 K. de Boer, "Stereophonic recording on Philips-Miller film," Philips Tech. Rev., 
vol. 6, pp. 80-84; 1941. 


Partly by reason of the low cost of the reproduction method, it is in 
principle possible that this ideal method of reproducing music will 
come within the reach of everyone for use in the home. An attraction 
of the cellophane films used for this method is that a music film 
with a playing time of one hour forms a reel no more than 18 cm 
in diameter (playing speed 32 cm per sec); see Fig. 9. This is 

Fig. 9. An illustration of the compactness of a recording of music on cello- 
phane film. Music that takes one hour to play can be recorded stereophonically 
on a film reel of the size of that shown in the illustration. For the same playing 
time (without stereophony!) 10 gramophone records of 25 cm diameter are re- 

due on the one hand to the extreme thinness of the cellophane 
film (40 microns) and on the other hand to the fact that it is 
possible to print on a 7-mm film two stereophonic sound tracks 
(thus in all four tracks). 

A piece of music of one hour can be reproduced without any of the 
interruptions that are unavoidable with the gramophone even when 
using an automatic record-changer. 


Picture-Sound Film 

Apart from the cinema there is a wide field of possibilities awaiting 
the "talkies." Such could be used on a large scale for entertainment 
in the home, for educational purposes in schools, for advertising, etc., 
provided they are cheap and of good quality. The 8-mm film, which 
would seem to lend itself best to this purpose, has not yet been widely 
used because it is still too expensive and owing to the limited resolving 
power of the common emulsions the pictures are not sharp enough; 
furthermore it does not leave any room for the sound track. Though 
the sound track is applied on a 16-mm film the quality of the sound re- 
production is not all one would desire. 

Now the metal-diazonium system as a copying material presents a 
situation that is more favorable in many respects: apart from the 
possibility of making a sound track even on an 8-mm film, on a 16- 
mm film a better sound quality can be obtained than has hitherto been 
possible, partly due to the high resolving power and partly by reason 
of the variability of the gamma. Both of these factors also help in im- 
proving the picture quality as is clearly noticeable on a contact print 
of a very fine-grained film, e.g. Isopan FF although a limit is set to 
the sharpness of the picture by the limited resolving power of the orig- 
inal film (about 55-75 lines/mm) and possibly of the optical system 
of the camera. Finally the film could be cheaper. 

The same considerations apply for the copying of standard 35-mm 
films on the new system. 

For playing the very thin cellophane films special projectors are re- 
quired. Figure 10 shows a model of a home cinema equipped with 
such a projector. If normal projectors are still to be used then the 
metal-diazonium system will have to be applied in or on a thicker 
carrier (say 130 microns thick), but then of course one loses the ad- 
vantage of the great compactness of film reels with a long playing time. 


Micro-documentation is a comparatively young branch of the 
technique of reproduction, but it seems that a surprising development 
may be expected in this very direction. 

The name itself already expresses its meaning: the recording of 
documents on a very small scale. The need for this may be due to 
various reasons. In some cases it is resorted to because the documents 
are so bulky or would become so bulky as to constitute a problem for 




their filing, storage, handling or transportation; examples are the cata- 
logs of large libraries, card index systems of large telephone exchanges 
or of the registers of births, deaths and marriages in large cities, etc. 
There are other cases where micro-documentation is applied because 
reproduction on the normal scale is too expensive, as for instance 

1 ! 

Fig. 10. Model of a home cinema fitted with a special projector (mounted in 
the cabinet) for cellophane film. The projected picture satisfies high require- 
ments regarding sharpness as well as gradation. 

the copying of publications by libraries, or for the air-mailing of docu- 
ments where weight is a big consideration, etc. 

Obviously the higher the resolving power of the material used for the 
photographic reproduction, the smaller the size of the reduced docu- 
ment. With the extremely high resolving power of our metal-diazo- 




Fig. 11. Three small sheets of metal-diazonium paper on which the complete 
contents of a book of 330 pp. have been reduced. Each page of the book is re- 
duced on the paper to a size of 5 X 7 mm. The reproduced book is quite legible 
with a simple reading apparatus. 

nium system it is possible to make a perfect record of a whole page of 
printing of the size of an 8 X llj^ in. periodical on an area of 0.6 by 
0.9 mm, the height of the letters being about 12 microns. 

It cannot be predicted whether one will ever go as far as such an 
extremely small size with the present stage of development of micro- 
documentation. For the present the sizes of, for instance, 5 X 7 or 
2X3 mm seem desirable. 


If one keeps to the larger dimensions of say 5X7 mm per page then 
the high resolving power of the metal-diazonium system is not utilized 
to its fullest extent, but even so this material will prove to be of great 
advantage owing to its low cost. Large tabular works, encyclopedias, 
etc., which can now practically only be consulted in libraries, could be 
reproduced on such a small scale on this inexpensive material as to be 
brought within the reach of anyone having occasion to read such works 
from time to time. Micro-reproductions can be read with a simple 
reading apparatus. The saving in volume is astonishing, even with 
the relatively large size of 5 X 7 mm, for a series of large books total- 
ing 10,000 pages can be reduced to a pocket-size booklet of 100 pp. 
Three such pages, compared with the original normal book, are shown 
in Fig. 11. 


As the last field of application for the metal-diazonium system we 
would mention that of macro-documentation, which has already be- 
come of common usage in the form of diazotype and blueprinting and 
photocopying directly on a legible scale. Owing to the nature of the 
documents to which this process is applied (drawings, specifications, 
etc.), for this purpose the metal-diazonium system would be em- 
ployed in paper. The paper is sensitized on both sides, so that it can 
be used on both sides for different copies; it does not curl up, and on 
account of its two-sided use it sometimes means a considerable saving 
in volume. Furthermore the image (a silver picture) gives a very 
rich contrast, with a pleasant tone (neutral grey), and is quite stable, 
properties which are often so lacking in diazotype and blueprinting 
processes. Of particular importance, however, is the possibility of 
making enlargements with metal-diazonium paper with exposure 
times that are practicable (for instance 1-5 seconds, depending upon 
the size). Thus we get a very efficient working method: wherever 
such is desired on account of frequent use, documents recorded 
on a micro-film can be enlarged again on metal-diazonium paper. 

Proposed Bylaw Amendment 

THE PROPOSED AMENDMENT to the Society's Bylaws that appears on 
the following pages was approved by the Board of Governors at its 
January 31st meeting, and is now ready for consideration by the voting 
members of the Society. At a Business Meeting now scheduled for the 
Opening Session of the 67th Convention in Chicago on April 24th, members 
will be asked to submit their comments or criticisms of the proposal and 
then vote on the adoption of the Amendment. If approved, the Amend- 
ment will take effect at once and will be published with the Constitution 
in the May issue of the JOURNAL. 

This is the second basic alteration of the rules governing the Society to 
be considered as a result of a three-year study of the Constitution, Bylaws 
and Administrative Practices. The first, developed early in 1949, produced 
the Constitutional Amendment, including the change of the Society's 
name, discussed at the Annual Business Meeting on October 10th in Holly- 
wood, and subsequently approved by the voting membership for adoption 
as of January 1, 1950. 

A special Committee on Revision of the Constitution and Bylaws was 
authorized by the Board of Governors and first appointed on January 23, 
1947, by then President, Loren L. Ryder. It has been continued by Presi- 
dent Earl I. Sponable and will now turn its attention to the task of prepar- 
ing a parallel revision of the Administrative Practices. 

The guiding principle in this work has been that of eliminating excess 
verbiage, together with a careful defining of committee procedures and re- 
sponsibilities, where necessary. Examples of the effect of this reappraisal 
are the recent changes in organization of certain engineering committees, 
adopted on recommendation of J. A. Maurer, former Engineering Vice-Presi- 
dent, and F. E. Carlson, Chairman of the Standards Committee. Greater 
responsibility for preparator3 r work on standards has been placed on mem- 
bers of the Standards Committee, which is now made up largely of chair- 
men of the several engineering committees. 

Members are urged to compare the following proposed Amendment with 
the present Bylaws, last published in the JOURNAL in April, 1949, p. 463. 
The recent Constitutional Amendment, including the change of name, last 
appeared, in the form as finally adopted, in the September, 1949, JOURNAL, 
p. 306. All voting members who can arrange to attend the April 24th 
Business Meeting are encouraged to do so and are invited to discuss freely 
the merits of the following proposal : 





Sec. 1. Membership of the Society shall 
consist of the following grades: Honorary 
members, Sustaining members, Fellows, 
Active members, Associate members and 
Student members. 

An Honorary member is one who has per- 
formed eminent service in the advance- 
ment of engineering in motion pictures, 
television, or allied arts. An Honorary 
member shall be entitled to vote and to 
hold any office in the Society. 

A Sustaining member is an individual, 
company, or corporation subscribing sub- 
stantially to the financial support of the 

A Fellow is one who shall be not less than 
thirty years of age and who shall by his 
proficiency and contributions have at- 
tained to an outstanding rank among engi- 
neers or executives of the motion picture 
or television industries. A Fellow shall be 
entitled to vote and to hold any office in 
the Society. 

An Active member is one who shall be not 
less than twenty-five years of age and shall 
be or shall have been either one or an 
equivalent combination of the following: 

(a) An engineer or scientist in motion 
picture, television or allied arts. As such 
he shall have performed and taken re- 
sponsibility for important engineering or 
scientific work in these arts and shall have 
been in the active practice of his profession 
for at least three years, or 

(b) A teacher of motion picture, tele- 
vision or allied subjects for at least six 
years in a school of recognized standing in 
which he shall have been conducting a 
major course in at least one of such fields, 

(c) A person who by invention or by 
contribution to the advancement of engi- 
neering or science in motion picture, tele- 
vision or allied arts, or to the technical 
literature thereof, has attained a standing 
equivalent to that required for Active 
membership in (a), or 

(d) An executive who for at least three 
years has had under his direction impor- 
tant engineering or responsible work in the 
motion picture, television or allied indus- 
tries and who is qualified for direct super- 

vision of the technical or scientific fea- 
tures of such activities. An Active member 
shall be entitled to vote and to hold any 
office in the Society. 

An Associate member is one who shall be 
not less than eighteen years of age, and 
shall be a person who is interested in the 
study of motion picture or television tech- 
nical problems or connected with the 
application of them. An Associate mem- 
ber is not privileged to vote, to hold office 
or to act as chairman of any committee, 
although he may serve upon any commit- 
tee to which he may be appointed; and, 
when so appointed, shall be entitled to the 
full voting privileges on action taken by 
the committee. 

A Student member is any person regis- 
tered as a student, graduate or under- 
graduate, in a college, university, or other- 
educational institution of like scholastic 
standing, who evidences interest in motion 
picture or television technology. Member- 
ship hi this grade shall not extend more 
than one year beyond the termination of 
the student status described above. A 
student member shall have the same privi- 
leges as an Associate member of the Soci- 

Sec. 2. All applications for membership 
or transfer should be made on blank forms 
provided for the purpose, and shall give a 
complete record of the applicant's educa- 
tion and experience. Honorary and Fel- 
low grades may not be applied for. 

Sec. 3. (a) Honorary membership may 
be granted upon recommendation of the 
Honorary Membership Committee when 
confirmed first by a three-fourths majority 
vote of those present at a meeting of the 
Board of Governors, and then by a four- 
fifths majority vote of all voting members 
present at any regular meeting or at a 
special meeting called as stated in the by- 
laws. An Honorary member shall be ex- 
empt from the payment of all dues. 

(b) Upon recommendation of the Fellow 
Award Committee, when confirmed by a 
three-fourths majority vote by those pres- 
ent at a meeting of the Board of Gover- 
nors, an Active member may be made a 

(c) An Applicant for Active membership 
shall give as references at least two mem- 



bers of the grade applied for or of a higher 
grade. Applicants shall be elected to 
membership by a three-fourths majority 
vote of the entire membership of the ap- 
propriate Admissions Committee. An 
applicant may appeal to the Board of 
Governors if not satisfied with the action 
of the Admissions Committee, in which 
case approval of at least three-fourths of 
those present at a meeting of the Board 
of Governors shall be required for election 
to membership or to change the action 
taken by the Admissions Committee. 

(d) An applicant for Associate member- 
ship shall give as reference one member of 
the Society, or two persons not members 
of the Society who are associated with the 
motion picture, television, or allied indus- 
try. Applicants shall be elected to mem- 
bership by approval of the Chairman of 
the appropriate Admissions Committee. 

(e) An applicant for Student member- 
ship shall be sponsored by a member of the 
Society, or by a member of the staff of the 
department of the institution he is attend- 
ing, this faculty member not necessarily 
being a member of the Society. Applicants 
shall be elected to membership by approval 
of the Chairman of the appropriate admis- 
sions committee. 

Sec. 4- Any member may be suspended 
or expelled for cause by a majority vote 
of the entire Board of Governors, provided 
he shall be given notice and a copy in writ- 
ing of the charges preferred against him, 
and shall be afforded opportunity to be 
heard ten days prior to such action. 



Sec. 1. An officer or governor shall be 
an Honorary member, Fellow, or an Ac- 
tive member. 

BYLAW 111 


Sec. 1. The Board of Governors shall 
transact the business of the Society in ac- 
cordance with the Constitution and By- 

Sec. 2. The Board of Governors may act 
on special resolutions between meetings, 
by letter ballot authorized by the Presi- 
dent. An affirmative vote from a majority 
of the total membership of the Board of 
Governors shall be required for approval 
of such resolutions. 

Sec. S. A quorum of ten members of the 

Board of Governors shall be present to 
vote on resolutions presented at any meet- 
ing. Unless otherwise specified, a majority 
vote of the Governors present shall con- 
stitute approval of a resolution. 

Sec. 4. A member of the Board of Gover- 
nors may not authorize an alternate to act 
or vote in his stead. 

Sec. 5. Vacancies in the offices or on the 
Board of Governors shall be filled by the 
Board of Governors until the annual elec- 
tions of the Society. 

Sec. 6. The Board of Governors, when 
filling vacancies in the offices or on the 
Board of Governors, shall endeavor to 
appoint persons who in the aggregate are 
representative of the various branches or 
organizations of the industries interested 
in the activities of the Society to the end 
that there shall be no substantial predom- 
inance upon the Board, as the result of its 
own action, of representatives of any one 
or more branches or organizations of such 

Sec. 7. The time and place of all except 
special meetings of the Board of Governors 
shall be determined by the Board of 

Sec. 8. Special Meetings of the Board of 
Governors shall be called by the President 
with the proviso that no meeting shall be 
called without at least seven days prior 
notice to all members of the Board by 
letter or telegram. Such a notice shall 
state the purpose of the meeting. 



Sec. 1. Special rules relating to the 
administration of the Society and known 
as Administrative Practices shall be es- 
tablished by the Board of Governors 
and shall be added to or revised as neces- 
sary to the efficient pursuit of the Society's 



Sec. 1. All committees, except as other- 
wise specified, shall be formed and ap- 
pointed in accordance with the Adminis- 
trative Practices as determined by the 
Board of Governors. 

Sec. 2. All committees, except as other- 
wise specified, shall be appointed to act 
for the term served by the officer charged 
with appointing the committees or until 
he terminates the appointment. 




-Sec. 3. Chairmen of the committees 
shall not be eligible to serve in such ca- 
pacity for more than two consecutive 

Sec. 4- Standing Committees of the 
Society to be appointed by the President 
and confirmed by the Board of Governors 
are as follows: 

Honorary Membership Committee 
Journal Award Committee 
Nominating Committee 
Progress Medal Award Committee 
Public Relations Committee 

' Samuel L. Warner Memorial Award 

Sec. 5. There shall be an Admissions 
Committee for each Section of the Society 
composed of a chairman and three mem- 
bers of which at least two shall be members 
of the Board of Governors. 

Sec. 6. There shall be a Fellow Award 
Committee composed of all the officers 
and section chairmen of the Society under 
the chairmanship of the Past-President. 
In case the chairmanship is vacated it shall 
be temporarily filled by appointment by 
the President. 


Sec. 1. The location and time of each 
meeting or convention of the Society shall 
be determined by the Board of Governors. 

Sec. 2. The grades of membership en- 
titled to vote are defined in Bylaw I. 

Sec. 3. A quorum of the Society shall 
consist in number of Ks of the total of 
those qualified to vote as listed in the 
Society's records at the close of the last 
fiscal year before the meeting. 

Sec. 4. The annual meeting shall be held 
during the fall convention. 

Sec. 5. Special meetings may be called 
by the President and upon the request of 
any three members of the Board of Gover- 
nors not including the President. 

Sec. 6. All members of the Society in any 
grade shall have the privilege of discussing 
technical material presented before the 
Society or its Sections. 


Sec. 1. The President shall preside at 
all business meetings of the Society and 
shall perform the duties pertaining to that 
office. As such he shall be the chief execu- 
tive of the Society, to whom all other offi- 
cers shall report. 

Sec. 2. In the absence of the President, 
the officer next in order as listed in Article 
V of the Constitution shall preside at 
meetings and perform the duties of the 

Sec. 3. The seven officers shall perform 
the duties separately enumerated below 
and those defined by the President: 

(a) The Executive Vice-President shall 
represent the President, and shall be re- 
sponsible for the supervision of the general 
affairs of the Society as directed by the 

The President and the Executive Vice- 
President shall not both reside in the geo- 
graphical area of the same Society Section, 
but one of these officers shall reside in the 
vicinity of the executive offices. Should 
the President or Executive Vice-President 
remove his residence to the same geo- 
graphical area of the United States as the 
other, the office of Executive Vice-Presi- 
dent shall immediately become vacant and 
a new Executive Vice-President shall be 
elected by the Board of Governors for the 
unexpired portion of the term. 

(b) The Engineering Vice-President 
shall appoint all technical committees. He 
shall be responsible for the general initia- 
tion, supervision, and co-ordination of the 
work of these committees. 

(c) The Editorial Vice-President shall be 
responsible for the publication of the 
Society's Journal and all other Society 

(d) The Financial Vice-President shall 
be responsible for the financial operations 
of the Society, and shall conduct them in 
accordance with budgets prepared by him 
and approved by the Board of Governors. 

(e) The Convention Vice-President 
shall be responsible for the national con- 
ventions of the Society. He shall arrange 
for at least one annual convention to be 
held in the fall of the year. 

Sec. 4- The Secretary shall keep a record 
of all meetings; and shall have the respon- 
sibility for the care and custody of records, 
and the seal of the Society. 

Sec. 5. The Treasurer shall have charge 
of the funds of the Society and disburse 
them as and when authorized by the Finan- 
cial Vice-President. He shall be bonded 
in an amount to be determined by the 
Board of Governors, and his bond shall 
be filed with the Secretary. 

Sec. 6. Each officer of the Society, upon 
the expiration of his term of office, shall 




transmit to his successor a memorandum 
outlining the duties and policies of his 



Sec. 1. All officers and governors shall be 
elected to their respective offices by a 
majority of ballots cast by voting members 
in the following manner: 

Nominations shall first be presented by 
a Nominating Committee appointed by 
the President, consisting of nine members, 
including a Chairman. The committee 
shall be made up of two Past-Presidents, 
three members of the Board of Governors 
not up for election, and four other voting 
members, not currently officers or gover- 
nors of the Society. Nominations shall 
be made by three-quarters affirmative 
vote of the total Nominating Committee. 

Not less than three months prior to the 
Annual Fall Meeting, the Board of Gov- 
ernors shall review the recommendations 
of the Nominating Committee, which shall 
have nominated suitable candidates for 
each vacancy. 

Such nominations shall be final unless 
any nominee is rejected by a three- 
quarters vote of the Board of Governors 
present and voting. The Secretary shall 
then notify these candidates of their 
nomination. From the list of acceptances, 
not more than three names for each va- 
cancy shall be selected by the Board of 
Governors and placed on a letter ballot. 
A blank space shall be provided on this 
letter ballot under each office, in which 
space the name of any voting member 
other than those suggested by the Board 
of Governors may be voted for. The bal- 
loting shall then take place. The ballot 
shall be enclosed with a blank envelope 
and a business reply envelope bearing the 
Secretary's address and a space for the 
member's name and address. One set of 
these shall be mailed to each voting mem- 
ber of the Society, not less than forty days 
in advance of the annual fall meeting. 

The voter shall then indicate on the 
ballot one choice for each vacancy, seal the 
ballot in the blank envelope, place this in 
the envelope addressed to the Secretary, 
sign his name and address on the latter, 
and mail it in accordance with the instruc- 
tions printed on the ballot. No marks of 
any kind except those above prescribed 
shall be placed upon the ballots or enve- 

lopes. Voting shall close seven days be- 
fore the opening session of the annual fall 

The sealed envelope shall be delivered 
by the Secretary to a Committee of Tellers 
appointed by the President at the annual 
fall convention. This committee shall 
then examine the return envelopes, open 
and count the ballots, and announce the 
results of the election. 

The newly-elected officers and governors 
of the Society shall take office on January 
1, following their election. 



Sec. 1. The annual dues shall be fifteen 
dollars ($15) for Fellows and Active mem- 
bers, ten dollars ($10) for Associate mem- 
bers, and five dollars ($5) for Student 
members, payable on or before January 1, 
of each year. Current or first year's dues 
for new members in any calendar year 
shall be at the full annual rate for those 
notified of election to membership on or 
before June 30; one half the annual rate 
for those notified of election to membership 
in the Society on or after July 1. 

Sec. 2. (a) Transfer of membership to a 
higher grade may be made at any time 
subject to the requirements for initial mem- 
bership in the higher grade. If the trans- 
fer is made on or before June 30, the an- 
nual dues of the higher grade are required. 
If the transfer is made on or after July 1, 
and the member's dues for the full year 
have been paid, one half of the annual dues 
of the higher grade is payable less one 
half the annual dues of the lower grade. 

(b) No credit shall be given for annual 
dues in a membership transfer from a 
higher to a lower grade, and such transfers 
shall take place on January 1, of each year. 

Sec. 3. Annual dues shall be paid in ad- 

Sec. 4- Failure to pay dues may be con- 
sidered just cause for suspension. 



Sec. 1. The Society shall publish a tech- 
nical magazine to consist of twelve 
monthly issues, in two volumes per year. 
The editorial policy of the Journal shall be 
based upon the provisions of the Constitu- 
tion and a copy of each issue shah 1 be sup- 
plied to each member in good standing 
mailed to his last address of record. 




Copies may be made available for sale at 
a price approved by the Board of Gover- 


Sec. 1. Sections of the Society may be 
authorized in any locality where the voting 
membership exceeds twenty. The geo- 
graphic boundaries of each Section shall 
be determined by the Board of Governors. 
Upon written petition for the authoriza- 
tion of a Section of the Society, signed by 
twenty or more voting members, the 
Board of Governors may grant such 


Sec. 2. All members of the Society of the 
Motion Picture and Television Engineers 
in good standing^ residing within the geo- 
graphic boundaries of any local Section 
shall be considered members of that Sec- 

Sec. 8. Should the enrolled voting mem- 
bership of a Section fall below twenty, or 
should the technical quality of the pre- 
sented papers fall below an acceptable 
level, or the average attendance at meet- 
ings not warrant the expense of maintain- 
ing that Section, the Board of Governors 
may cancel its authorization. 


Sec. 4- The officers of each Section shall 
be a Chairman and a Secretary-Treasurer. 
The Section chairmen shall be ex-officio 
members of the Board of Governors and 
shall continue in such positions for the 
duration of their terms as chairmen of the 
local Sections. Each Section officer shall 
hold office for one year, or until his suc- 
cessor is chosen. 


Sec. 5. The Board of Managers shall con- 
sist of the Section Chairman, the Section 
Past-Chairman, the Section Secretary- 
Treasurer, and six voting members. Each 
manager of a Section shall hold office for 
two years. Vacancies shall be filled by 
appointment by the Board of Managers 
until the annual election of the Section. 


Sec. 6. The officers and managers of a 
Section shall be voting members of the 
Society. All officers and managers shall 
be elected to their respective offices by a 

majority of ballots cast by the voting 
members residing in the geographical area 
of the Section. Not less than three 
months prior to the annual fall convention 
of the Society, nominations shall be pre- 
sented to the Board of Managers of the 
Section by a Nominating Committee ap- 
pointed by the Chairman of the Section, 
consisting of seven members, including a 
chairman. The committee shall be com- 
posed of the present Chairman, the Past- 
Chairman, two other members of the 
Board of Managers not up for election, and 
three other voting members of the Section 
not currently officers or managers of the 
Section. Nominations shall be made by 
a three-quarters affirmative vote of the 
total Nominating Committee. Such nom- 
inations shall be final, unless any nominee 
is rejected by a three-quarters vote of the 
Board of Managers, and in the event of 
such rejection the Board of Managers will 
make its own nomination. 

The Chairman of the Section shall then 
notify the candidates of their nomination. 
From the list of acceptances, not more than 
three names for each vacancy shall be 
selected by the Board of Managers and 
placed on a letter ballot. A blank space 
shall be provided on this letter ballot 
under each office, in which space the name 
of any voting member other than those 
suggested by the Board of Managers may 
be voted for. The balloting shall then 
take place. The ballot shall be enclosed 
with a blank envelope and a business reply 
envelope bearing the local Secretary- 
Treasurer's address and a space for the 
member's name and address. One of these 
shall be mailed to each voting member of 
the Society residing in the geographical 
area covered by the Section, not less than 
forty days in advance of the annual fall 

The voter shall then indicate on the 
ballot one choice for each office, seal the 
ballot in the blank envelope, place this in 
the envelope addressed to the Secretary- 
Treasurer, sign his name and address on 
the latter, and mail it in accordance with 
the instructions printed on the ballot. 
No marks of any kind except those above 
prescribed shall be placed upon the ballots 
or envelopes. Voting shall close seven 
days before the opening session of the 
annual fall convention. The sealed enve- 
lopes shall be delivered by the Secretary- 
Treasurer to his Board of Managers at a 




duly called meeting. The Board of Man- 
agers shall then examine the returned enve- 
lopes, open and count the ballots, and an- 
nounce the results of the election. 

The newly-elected officers and managers 
shall take office on January 1, following 
their election. 


Sec. 7. The business of a Section shall be 
conducted by the Board of Managers. 


Sec. 8. (a) At the beginning of each fiscal 
year, the Secretary-Treasurer of each sec- 
tion shall submit to the Board of Gover- 
nors of the Society a budget of expenses 
for the year. 

(b) The Treasurer of the Society shall 
deposit with each Section Secretary- 
Treasurer a sum of money for current ex- 
penses, the amount to be fixed by the 
Board of Governors. 

(c) The Secretary-Treasurer of each 
Section shall send to the Treasurer of the 
Society, quarterly or on demand, an item- 
ized account of all expenditures incurred 
during the preceding period. 

(d) Expenses other than those enu- 
merated in the budget, as approved by the 
Board of Governors of the Society, shall 
not be payable from the general funds of 
the Society without express permission 
from the Board of Governors. 

(e) The Section Board of Managers 
shall defray all expenses of the Section not 
provided for by the Board of Governors, 
from funds raised locally. 

(f) The Secretary of the Society shall, 
unless otherwise arranged, supply to each 
Section all stationery and printing neces- 
sary for the conduct of its business. 


Sec. 9. The regular meetings of a Section 
shall be held in such places and at such 
hours as the Board of Managers may desig- 
nate. The Secretary-Treasurer of each 
Section shall forward to the Secretary of 
the Society, not later than five days after 
a meeting of a Section, a statement of the 
attendance and of the business transacted. 


Sec. 10. Sections shall abide by the 
Constitution and Bylaws of the Society 
and conform to the regulations of the 
Board of Governors. The conduct of Sec- 
tions shall always be in conformity with 

the general policy of the Society as fixed 
by the Board of Governors. 



Sec. 1. Student Chapters of the Society 
may be authorized in any college, univer- 
sity, or technical institute of collegiate 
standing. Upon written petition for the 
authorization of a Student Chapter, signed 
by twelve or more Society members, or 
applicants for Society membership, and 
the Faculty Adviser, the Board of Gover- 
nors may grant such authorization. 


Sec. 2. All members of the Society in 
good standing who are attending the desig- 
nated educational institution shall be 
eligible for membership in the Student 
Chapter, and when so enrolled they shall 
be entitled to all privileges that such Stu- 
dent Chapter may, under the Constitution 
and Bylaws, provide. 

Sec. 3. Should the membership of the 
Student Chapter fall below ten, or the 
average attendance at meetings not war- 
rant the expense of maintaining the 
organization, the Board of Governors may 
cancel its authorization. 


Sec. 4- The officers of each Student 
Chapter shall be a Chairman and a 
Secretary-Treasurer. Each Chapter officer 
shall hold office for one year, or until his 
successor is chosen. Where possible, 
officers shall be chosen in May to take 
office at the beginning of the following 
school year. The procedure for holding 
elections shall be prescribed in Administra- 
tive Practices. 


Sec. 5. A member of the faculty of the 
same educational institution shall be 
designated by the Board of Governors as 
Faculty Adviser. It shall be his duty to 
advise the officers on the conduct of the 
Chapter and to approve all reports to the 
Secretary and the Treasurer of the Society. 


Sec. 6. The Treasurer of the Society shall 
deposit with each Chapter Secretary- 
Treasurer a sum of money, the amount to 
be fixed by the Board of Governors. The 
Secretary-Treasurer of the Chapter shall 
send to the Treasurer of the Society at the 



end of each school year or on demand an 
itemized account of all expenditures in- 


Sec. 7. The Chapter shall hold at least 
four meetings per year. The Secretary- 
Treasurer shall forward to the Secretary 
of the Society at the end of each school 
year a report of the meetings for that year, 
giving the subject, speaker, and approxi- 
mate attendance for each meeting. 



Sec. 1. Proposed amendments to these 
Bylaws may be initiated by the Board of 
Governors or by a recommendation to 
the Board of Governors signed by ten 
voting members. Proposed amendments 

may be approved at any regular meeting 
of the Society at which a quorum is present, 
by the affirmative vote of two-thirds of 
the members present and eligible to vote 
thereon. Such proposed amendments 
shall have been published in the Journal 
of the Society, in the issue next preceding 
the date of the stated business meeting of 
the Society at which the amendment or 
amendments are to be acted upon. 

Sec. 2. In the event that no quorum of 
the voting members is present at the time 
of the meeting referred to in Sec. 1, the 
amendment or amendments shall be re- 
ferred for action to the Board of Gover- 
nors. The proposed amendment or amend- 
ments then become a part of the Bylaws 
upon receiving the affirmative vote of 
three-quarters of the entire membership 
of the Board of Governors. 

16-Mm Sound Service Test Film 

USERS AND MANUFACTURERS of 16-mm sound projectors have long 
had available a comprehensive group of sound test films produced 
and sold by the Society and the Motion Picture Research Council. 
Although distribution of these films was not restricted, but on the 
contrary has been encouraged widely, the order of technical accu- 
racy demanded by the specifications under which they were made 
forced production costs up so high that their use was, in effect, 
restricted to professional applications. 

Realizing that there was a tremendous need for a more practical 
type of sound test film, combining on one reel several tests having 
an order of accuracy commensurate with the test equipment and 
tools in use by projector service men, the Society set about de- 
veloping such a film. The result is the recently announced 16-Mm 
Sound Service Test Film. It includes technical test sections that 
check the lateral position of the film as it passes the sound scanning 
beam, the focus of the sound reproducer optical system and the over- 
all frequency-response characteristic of the projector. There are 
also three music samples, and one section of dialogue that provide 
a critical over-all listening test. Together, these several sections 
will show whether or not poor sound performance is the fault of the 
projector, and if so, the nature of the trouble being encountered. 

Individual projector owners, schools, churches and service shops 
may now perform a specific series of tests, and by following the in- 
structions that appear on the screen, interpret the test results in 
terms of repairs or adjustments needed to put the equipment back 
in good order. 

TITLE MUSIC. The title music the sound system is correctly adjusted, 

is a specially recorded orchestral section this section, particularly in the ascend- 

selected as a subjective test of the fre- ing xylophone runs and the high-pitched 

quency range, high- and low-frequency bell tones, should be clear, crisp and 

balance, and distortion introduced by full, without harshness and without 

the 16-mm projector being tested. If quaver. 




This section is for use in making a quick 
focus adjustment of the sound optical 
system and contains 15 ft of square 
wave, 5,000-cycle track. Maximum 
loudness of the tone from the loud- 
speaker is a sufficient indication of cor- 
rect focus. If correct adjustment can- 
not be attained in the length of time 
provided using the normal focus con- 
trol, it is recommended that a careful 
shop adjustment be made using Service 
Type Sound Focusing Test Film, Z22.42, 
which is 100 ft in length and is sup- 
plied with complete instructions for use. 

BUZZ TRACK. When this section 
is run on the projector in correct ad- 
justment, that is when the sound-track 
scanning light beam is properly posi- 
tioned and the film does not weave from 
side to side, no tone should be heard 
from the loudspeaker. If the scanning 
beam is too near the edge of the film, a 
1000-cycle tone will be audible; or if it 
is too far from the edge a 300-cycle tone 
will be heard. Adjusting the guides 
that position the film laterally should 
eliminate both tones. The tones will be 
heard alternately if the film is weaving 
from one side to the other, and they will 
both be heard at once if the length of 
the scanning beam is too great. To 
remedy either of the last two faults, a 
shop adjustment should be made using 
the American Standard Buzz Track 
Film, Z22.57. 

This section is a subjective test which 
permits the listener to make a quick 
aural evaluation of the frequency re- 
sponse of the projector and its loud- 
speaker. There are twelve different 
frequencies ranging from 50 through 
6,000 cycles, each of which runs for 
about ten seconds. A 400-cycle tone 
precedes the 50-cycle section and also 
follows the 6,000-cycle section, as a vol- 
ume-level check. 

During reproduction of the first sec- 
tion, the tone control should be set at 

normal and the volume control ad- 
justed for a comfortable listening level. 
All tones that follow should be clearly 
audible and none should be uncomfort- 
ably loud. If any of the foregoing con- 
ditions is not met, the projector is not 
operating in a satisfactory manner. 

Assuming the buzz track and sound 
focusing section have indicated correct 
optical adjustment, failure to reproduce 
the multi-frequency section properly 
indicates trouble in the amplifier or 
associated loudspeaker system. Adjust- 
ment and repair of these components 
require the use of precision instruments 
and a carefully calibrated multi-fre- 
quency test film made in accordance 
with American Standard, Z22.44. 

DIALOGUE. This section is in- 
cluded as an example of good dialogue 
recording. Sibilants should be clear 
and crisp, while low tones should be 
full. With the loudspeaker placed near 
or directly behind the screen there 
should be a definite feeling of "pres- 

PIANO MUSIC. When this sec- 
tion is correctly reproduced it should 
give the reverberant effect of being 
played in a "live" room. The full tonal 
range should reproduce well, particu- 
larly without appreciable waver of sus- 
tained notes. Failure to meet this lat- 
ter requirement indicates varying rate 
of motion as the film passes the sound 
scanning beam and the projector should 
be checked using American Standard 
3,000-Cycle Flutter Test Film, Z22.43. 

section should give the effect of being 
played in a large auditorium having a 
slight reverberation characteristic. The 
orchestra should sound well balanced 
and the volume range should be handled 
by the projector without the necessity 
for changing the adjustment of the vol- 
ume controls. High-level passages 
should reproduce without distortion 
while the low-level portions should be 
clearly audible. 

67th Semiannual Convention 

To bring yourself and your company up to date on the technical 
trends in motion pictures and television, you can do nothing better 
than attend the 67th Convention of the Society that will be held at 
The Drake hotel in Chicago from April 24 to 28. 

In ten technical sessions, to be held two each day from Monday 
through Friday, you will hear papers on many subjects including, 
among others, projection arc lamps, production techniques for tele- 
vision studios, high-speed photography and color. You will also be 
invited to take part in the informal discussions that follow the pre- 
sentation of all papers. 

HOTEL RESERVATIONS: Rooms have been set aside for all members and their 
friends who wish to attend. Mr. John Gorte, Office Manager, The Drake, Chi- 
cago 11, 111., will confirm your accommodations promptly if you will mail him the 
room reservation card you received during the first week of March. If you haven't 
mailed it, do so without delay to be certain of your accommodations in advance. 

TRAVEL: Arrange for your train or plane reservations both to and from Chicago 
well ahead of time, because week-end travel is particularly heavy in and around 
Chicago. Your local travel agent will be glad to schedule your trip at least one 
month in advance. 

CONVENTION REGISTRATION: Registration will begin at 9:30 Monday 
morning, April 24, in the hotel's French Room Foyer. The Registration Desk 
will remain open during all technical sessions, with luncheon and banquet tickets 
available from either E. R. Geib or W. C. Kunzmann. 

LUNCHEON: A prominent speaker will address the 'Get-Together' Luncheon, 
scheduled for 12: 30 Monday, April 24, in the hotel's Gold Coast Room. To assure 
seating, tickets should be purchased in advance, and table reservations must also 
be made with Mr. Kunzmann as early as possible. If you have special require- 
ments, it would be well for you to write him at: National Carbon Division, Box 
6087, Cleveland, 1, Ohio. 

BUSINESS SESSION: Immediately following the luncheon, all the Society's 
Voting Members will be asked to attend an official Business Meeting of the Society 
in the Grand Ballroom to discuss and vote on the proposed new Society Bylaws 
that are published on p. 367 of this JOURNAL. This is the only business scheduled 
for the meeting, and since voting doubtless will be concluded promptly, the first 
technical papers of the Convention will be presented without delay. 

COCKTAIL HOUR AND BANQUET: Wednesday evening is the tune for frivolity. 
Bill Kunzmann will serve as host to all members and guests at the Cocktail Party 
which begins at 6:45 in the French Room and the informal banquet scheduled 
to start at 8:00 o'clock in the Gold Coast Room. There will be music, dancing, 


entertainment and fun for all. Tickets for the banquet should be purchased 
early. Table reservations must be made with Mr. Kunzmann at the Registration 
Desk before Wednesday so that members may be assured of places for themselves 
and their guests. 

LADIES' ACTIVITIES: For the wives and guests of members who plan to be in 
Chicago during convention week, Mrs. G. W. Colburn, Convention Hostess, and 
the members of the Ladies' Reception Committee have arranged a program that 
will prove to be both interesting and entertaining. The details of these plans will 
be discussed at the Ladies' Registration Headquarters in Parlor H by members of 
the Ladies' Reception Committee, who will be on hand daily. 

RECREATION: Members and their guests who register will receive complimen- 
tary passes for the week to several de luxe motion picture theaters in Chicago. 
Chicago has much to offer in the way of plates of historical interest and other at- 
tractions that will be described at the Registration Desk. 

PLAN TO ATTEND: Bring your friends and make a successful week of this - 
the 67th Semiannual Convention. 

Society Announcements 

Student Chapter 

The Student Chapter of the Society at New York University was formally 
established by action of the Board of Governors on January 31, 1950. This is 
the second Student Chapter. The first was organized at the University of South- 
ern California and is presently under the Chairmanship of Algernon Walker. 

Formation of a New York University Chapter was first proposed to the Society 
on November 9, 1949, by William F. Boden, a student in the Motion Picture 
Dept., representing 15 applicants for Student membership. Subsequently, they 
petitioned the Board of Governors for authorization to form a Student Chapter 
within the limitations established under the Bylaws of the Society. 

After duly considering the petition, which had been endorsed by Professor 
Robert Gessner, Chairman of NYU's Motion Picture Dept., the Board of Gover- 
nors voted to authorize formation of the Chapter, and extended to all Chapter 
members its enthusiastic welcome. 

The Chapter Officers are: William F. Boden, Chairman; Gerald I. Rosenfeld, 
Secretary-Treasurer; and Professor Gessner, Faculty Adviser. 

New Sustaining Members 

The Society is always pleased to welcome additions to its list of Sustaining 
Members. Each new one who joins provides additional financial support needed 
to help carry out the many projects assigned to engineering committees and to 
continue the ambitious publications program being undertaken. In addition, the 
continued increase of sustaining support is a measure of proof that the Society's 
contributions to the fields that lie within the area of its technical scope are both 
valid and effective. 

The Society is pleased to acknowledge these Sustaining Memberships, received 
during the past month: Blumenfeld Theaters, DuArt Film Laboratories, Inc., 
Hallen Qorp,, National Screen Service and Producers Service Co. 

Film Decomposition Tests 

In the paper that begins on p. 268 of this JOURNAL, Cummings 
Button and Silfin point out the serious losses that may result from 
decomposition in storage of cellulose nitrate motion picture film. 
Much of this potential loss could no doubt be avoided if a reliable 
test for future storage life were applied to films now being stored for 
commercial and archival purposes. No such test is now used regu- 
larly in the United States, but recently the attention of film librarians 
has been directed to two such tests developed in England, under 
the auspices of the British Government's Chemical Research and 
Development Establishment and the Department of the Govern- 
ment Chemist. In Report No. 2/R/48, "The Surveillance of 
Cinematograph Record Film During Storage/' by G. L. Hutchison, 
L. Ellis, and S. A. Ashmore, the authors described a rather extensive 
investigation of the stability of film in storage and outlined the 
details of two test procedures which they have found useful. Fur- 
ther information concerning the use of these tests and the results 
obtained may be secured from the British Film Institute, 164 
Shaftesbury, London, England. This is a government department 
similar to the U.S. National Archives. 

So that the SMPTE's Preservation of Film Committee may have 
the benefit of a number of points of view concerning both the value 
of such tests generally and the advisability of further committee 
study in this direction, a brief summary of this British Report and 
an outline of the test methods are described immediately below. 
Comments and recommendations should be addressed to James W. 
Cummings, Chairman, SMPTE Preservation of Film Committee, 
The National Archives, Washington 25, D.C. 


"The deterioration of nitrocellulose useful life of the film, and as such film 

base cinematograph film on pro- cannot be duplicated it represents a 

longed storage is brought about by a total loss of record, 

slow but progressive decomposition "Two tests, based on methods of 

of the nitrocellulose. The changes known value in the examination of 

occurring are complex but it seems nitrocellulose explosives, have been 

clear that the gelatine on the film acts developed whereby it is possible to 

as a stabilizer and accordingly suffers anticipate the end of the useful life of 

deterioration which involves loss of a film. The results of these tests 

the contained silver image at about allow sufficient time for a film to be 

the same time that the film becomes duplicated while still in good physical 

sticky. This stage is the end of the condition. 


"Using the two tests referred to in 
the above paragraph a scheme of 
surveillance and sentencing of stored 
films has been devised. 

"Since the useful life of a film is 
well ended before conditions favor- 
able to spontaneous inflammation 
arise, it is clear that danger from this 
source can now be avoided. 

"The results of this investigation 

are clearly applicable in principle to 
all stored cinematograph film having 
a cellulose nitrate base." 

Two recommended tests for pre- 
dicting the future condition of ni- 
trate film in storage are the Alizarin 
Red Heat Test and the Micro- 
Crucible Test. These are described 
briefly below: 


In this test a small punching of the 
film of approximately 6 mm diam- 
eter and weighing 7 mg is heated in a 
glass tube in which is suspended an 
alizarin red test paper moistened with 
a solution of glycerine in water. The 
tube is heated at a temperature of 
134 C and the time noted for the 
development of acid vapors as indi- 
cated by a color change in the test 

1. Apparatus 

(a) Glass tube closed at one end 
and approximately 90 mm in length 
and 9 mm internal diameter is fitted 
with a stopper 60 mm in length and 
of such a diameter that when covered 
with one thickness of test papers 
makes a close sliding fit in the tube. 
45 mm of the stopper should then be 
in the tube. 

(b) A cylindrical double-walled 
copper air bath, 100 mm deep and 
100 mm in diameter. The metal lid 
is lagged with asbestos composition 
8 mm thick and contains a central 
hole to take a thermometer and six 
holes, of 12-mm diameter, suitably 
arranged to take six testing tubes. 
The outer jacket of the bath is fitted 
with a reflux condenser. 

(c) A supply of rubber rings to 
give convenient support to the tubes 
in the air bath. 

(d) A supply of large size filter 
papers (Whatman, No, 2), 


2. Preparation of the Test Paper 
This is conveniently prepared by 

impregnating a sheet of filter paper 
with a 0.1% solution of alizarin red 
indicator in water to which has been 
added 2 ml of 2 N ammonia per 100 
ml. The solution is allowed to drain 
off and the edges of the paper are sub- 
sequently discarded. The paper may 
then be air dried but should be fur- 
ther heated for 10 min in the steam 
oven before use, to drive off any 
traces of free ammonia. 

3. Method 

The outer jacket of the bath 'is 
charged with about 20 ml of pure 
xylene and heat is applied by means 
of a small gas burner. When tem- 
perature conditions are steady the 
thermometer should read 134 C. 

In the meantime six tubes, as de- 
scribed at (a) under Apparatus, are 
prepared for test. Into each is placed 
a punching of the film under test, the 
punching being 6 mm in diameter and 
weighing about 7 mg. A strip of test 
paper is cut to such a width that 
when wrapped around the stopper the 
whole of the latter is effectively cov- 
ered without any overlap which 
would spoil the snug fit in the tube. 
The test paper is then moistened with 
a 50% solution of glycerine in water. 
Each tube is fitted with a rubber 
ring, the position of which is adjusted 
so that the part of the tube contain- 

ing the test paper is wholly outside 
the bath. 

With all tubes in position in the 
bath the temperature and time are 
noted. The tubes are kept under 
constant observation, if necessary for 
rather more than one hour. 

The color of the test paper pre- 
pared as described is maroon, which, 
in the presence of acid vapors, is 
either bleached or becomes a pale 
yellow. The change is well marked 
and readily observed: for the pur- 

poses of this test, the time in minutes 
required for a positive result is taken 
when the lower edge of the paper is 
bleached or changed in color to a dis- 
tance of approximately J in. 

Note 1. The apparatus described 
can suitably be increased in size so 
that more than six samples can be 
tested at one time. 

Note 2. If a tube is removed after 
test while the other tubes are in posi- 
tion, the vacant hole must be closed 
with a cork or bung. 


This test involves the determina- 
tion of the loss in weight of a punch- 
ing of cinema film when heated in a 
small porcelain crucible in a venti- 
lated oven maintained at 100 C. 

The crucibles are of 1 ml capacity 
and are obtainable from Royal Wor- 
cester Porcelain Co., or indirectly 
through laboratory supply firms. 

A disc of film of approximately 
0.25-in. diameter and weight about 7 
mg is punched out of the film in 
question and transferred to the 

Alizarin Red Heat Test 

crucible weighed to the nearest 0.01 
mg. The combined weight of cru- 
cible and film is then determined ac- 
curately, also to 0.01 mg, and the 
weight of film found by difference. 

The crucible and film are then 
heated in a ventilated oven main- 
tained at 100 C and the combined 
weight determined at 168 hr and 300 
hr. The loss in weight is then calcu- 
lated as a percentage on the original 
weight of the film. 


(after 168 hr) 

60 min and over 

Under 60 min but not under 30 min 

Under 30 min but not under 10 min Under 10% 

Retest after: 


6 months 

Under 10 min 

10% and over Copy and destroy 
Copy and destroy 

Meetings of Other Societies 

Institute of Radio Engineers, Cincinnati Section, Spring Technical Conference on 

Television, April 29, Cincinnati, Ohio 

Institute of Radio Engineers, Technical Conference, May 3-5, Dayton, Ohio 
Armed Forces Communications Assn., Annual Meeting, 

May 12, New York, and Long Island City 
May 13, Fort Monmouth, N.J. 

Acoustical Society of America, Spring Meeting, June 22-24, State College, Pa. 
Illuminating Engineering Society, National Technical Conference, 

August 21-25, Pasadena, Calif. 

Engineering Committees 

Theater Television 

On February 14, the Society transmitted to the Federal Com- 
munications Commission a formal reply to the Commission's Public 
Notice of January 11 which outlined its plans for a hearing on alloca- 
tions and rule making for a theater television service. The ten 
points at issue in the present controversy, as stated by the Commis- 
sion, were published on p. 237 of the February JOURNAL. The 
Society's reply read as follows: 

"Before the 

Washington 25, B.C. 

"In the Matter of 

Allocation of Frequencies and Promulgation of Rules and Docket No. 9552 

Regulations for a Theater Television Service 


"In accordance with Provision 7 of Public Notice No. 45051 of the Federal 
Communications Commission, dated January 11, 1950, and dealing with a hearing 
on Theater Television, the Society of Motion Picture and Television Engineers 
hereby files its written appearance in this proceeding, and gives notice that it will 
appear through its appointed representatives. They will present evidence on the 
issues specified in the aforementioned notice of hearing issued by the Commission 
upon such data, and at such time and place as may be selected by the Commission 

"Submitted for the Society of Motion Picture and Television Engineers by 

(Signed) E. I. SPONABUS 


The commercial success over the last ten years of a number of different processes 
for motion picture release printing in color has been a substantial boon to the 
industry but has simultaneously introduced several very practical problems that 
will undoubtedly remain with us for some time to come. Limited familiarity with 
the language of color as a general branch of physics or as a highly specialized 
branch of motion picture engineering is an obstacle to many engineers and techni- 
cians who encounter these color systems in their daily work. In research or film 
processing laboratories where the nomenclature is better understood, the lack of 
uniform methods of measurement or specification complicates the already difficult 
problem of drawing significant comparisons between competitive processes. 
Commercial applications which involve the use of more than one manufacturer's 
product at succeeding steps in the complete process are made exceedingly complex 
and for the same reason the recording, duplicating and reproducing of photo- 
graphic sound tracks on color films are made many times more difficult than is the 
case with familiar black-and-white emulsions. 


To shed some additional light on this latter problem, a Color Subcommittee, 
under the Chairmanship of Lloyd T. Goldsmith, has prepared a tabulation of the 
characteristics of sound tracks produced on commercial 35- and 16-mm color 
print processes. The table appears on p. 377 of this JOURNAL. 

Not only manufacturers and users of sound reproducing equipment, but also 
film laboratory technicians will find this table of real value. Reprint copies on 
heavy paper have been prepared and will be supplied free of charge by the Society 
to all who wish them. Requests for copies or comments on the tabulation should 
be addressed to Bill Deacy at Society Headquarters. 

An early issue of the JOURNAL will carry a more extensive contribution to the 
available literature on color motion pictures. This will be in the form of a report 
on the "Principles of Color Sensitometry," prepared by the Color Sensitometry 
Subcommittee under the Chairmanship of Carl F. J. Overhage. This report, 
which has been in work for nearly a year, is now completed and in manuscript form 
it amounts to nearly 150 pp. 

Film Dimensions 

Methods of producing 16-mm release prints in large quantity through the use 
of 32-mm perforated films have been adopted widely in the United States during 
the last decade. Experience gained over this period has resulted in the gradual 
development of uniform practices in preparation of negative picture and sound 
material, in printing the release positives and in slitting after processing to produce 
the 16-mm prints for projection. Work was begun in 1948 on standards for the 
special films thus used. Formal proposals for the standardization of two 32-mm 
films and one 35-mm film, 32-mm perforated, were published for a period of trial 
and criticism in the February, 1949, JOURNAL. Shortly after publication, a ques- 
tion arose concerning commercial slitting tolerances of 32-mm raw stock but on 
further investigation raw stock was found to be within the limits published in the 
proposed standards. As a result objections to the original proposals were with- 
drawn and two of them, previously approved by the Standards Committee, have 
now been forwarded to the ASA Sectional Committee on Motion Pictures, Z22. 
The third, covering the Dimensions of 35-mm film with 32-mm perforations, had 
not previously been sent to the Standards Committee, so is now out for their con- 
sideration. When the action of the Standards Committee is completed, this pro- 
posal will also be submitted to the Sectional Committee. 

Book Review 

Introduction to Theoretical and Experimental Optics, by Joseph 

Published (1949) by John Wiley and Sons, 440 Fourth Ave., New York 16. 
429 pp. + 6 pp. appendix + 4 pp. "Answers to Problems" + 4 pp. bibliography + 
10 pp. index. 44 illus. 5% X 8% in. Price $6.50. 

Publication of a new textbook of optics is a rather rare event these days when 
general attention is diverted to the more spectacular topics which abound in 
nuclear physics and electronics. A good general reason exists, therefore, to wel- 
come this book. This is, however, not the only reason, as the book has its own 


merits and it constitutes a valuable contribution to the fundamental optical litera- 

The purpose of this book is to give the college student a working knowledge of 
the extended field of optics. Accordingly, the text covers in its four parts geo- 
metrical optics, physical optics, radiation and spectra, and a series of laboratory 
experiments under the heading of experimental optics. The first three parts are 
divided into 21 chapters accompanied by problems. 

The presentation of the subject matter is well balanced and necessarily con- 
densed, for it would not be possible to cover all the branches of optics and some 
closely related subjects in one handy volume. As was noted by the author and 
the publishers, some subjects (X rays, photographic optics, and ophthalmic 
lenses) are treated in greater detail than is customary in general textbooks of 
optics. This deviation from the usual is not objectionable to the reviewer. Al- 
though the reviewer cannot offer definitive rules as to what should be included and 
what may be excluded from a college textbook, he feels that no textbook should 
ignore the phenomena and objects with which we are in practically everyday con- 
tact. Indeed, it is rather discouraging to meet students who are conversant with 
the Kerr, Zeeman, Raman and other "effects," but have very little to say about 
their own spectacles or their photographic lenses. Let us hope that Prof. Valasek's 
students will be well acquainted with both laboratory and everyday aspects of 

Following this thought, the reviewer would have been pleased had the author 
at least briefly touched also the following subjects: medical and industrial radiol- 
ogy, biological effects of radiation, condensing and projection systems, anti-reflec- 
tion films, interference filters, and phase-contrast microscopy. It is particularly 
difficult to justify the omission of the last three subjects which are now in the 
limelight of optical engineering. 

The reviewer is disturbed by the fact that no recognition is given in the book 
to the current standardization efforts in the field of optics. Thus, while the term 
"equivalent focal length" has been widely used and recently sanctioned as stand- 
ard by the American Standards Association (Z38.4.21 1948), it is not even men- 
tioned in the book. It is beside the point whether or not the term is satisfactory 
(the reviewer is of the opinion that the "equivalent" is superfluous and mislead- 
ing), as perhaps no terminology can satisfy everybody. The fact is that a strong 
demand exists for standardization of terms, definitions, and procedures in optics, 
and that a serious effort is being made, under the sponsorship of the Optical 
Society of America and of other organizations, to satisfy the demand. This effort 
will be in vain if our educational institutions do not teach the younger generation 
to appreciate standardization and to adhere to it. It will also be disheartening 
to the younger generation to discover, for example, that, while knowing the H-D 
speed, the Schneider number, the DIN system, and the Weston rating, they know 
nothing about the American Standard Speed and the American Exposure Index 
(Z38.2.1 1947), which are not mentioned in this book. 

The book is not intended as an engineering manual. Still it may be very useful 
as a source of basic information to any engineering or research group concerned 
with optical problems. The well-selected bibliography at the end of the book 
adds considerably to its reference value. 

Bausch & Lomb Optical Co. 
Rochester 2, N.Y. 

Letter to the Editor 

In 'Twenty-Lens High-Speed Camera" by Charles W. Wyckoff, in the Novem- 
ber, 1949, JOURNAL, some acknowledgments were not mentioned. The design and 
construction of the twenty-lens camera was carried out undor the direction of Mr. 
Nrwt'll T. Partch who was then located at the David Taylor Model Basin. The 
camera itself was built at the Naval Observatory. The optical calculations were 
done by Dr. A. I. Mahan of the Naval Ordnance Laboratory. Mr. Wyckoff 
served as a consultant on the mechanical and electrical elements of the camera. 
Figures 6, 7 and 8 appearing in Mr. Wyckoff 's paper were the result of some of the 
previously mentioned optical calculations. The lens chosen for this camera was 
not a standard Navy lens as suggested by Mr. Wyckoff. It was a simple achro- 
matic doublet, whose characteristics were all carefully evaluated so that its per- 
formance would be known before inserting it in such a camera. Such achromatic 
doublets of the speed used here cannot be used at a field angle of 13 with good 
results. Nevertheless, this choice was made rather early to keep the cost of the 
first model of the camera down, with the result that the image quality deteriorated 
at the edge of the field even when the lens was stationary. 


U. S. Naval Ordnance Laboratory 
White Oak, Silver Spring 19, Md. 

Section Meetings 

Atlantic Coast 

"High-Speed Motion Pictures" will be the subject of the Atlantic Coast Section 
Meeting, scheduled for 7:30 P.M., Wednesday, March 22, in the Western Union 
Auditorium at 60 Hudson St., New York City. John H. Waddell, Chairman 
of the Society's High-Speed Photography Committee, is to be a speaker. 

Dr. Dirk Reuyl, Ballistics Research Laboratory, Aberdeen Proving Ground, 
Aberdeen, Md., will speak on "Optical Instrumentation for Guided Missiles." 


Two papers originally reported in February as being planned for presentation 
at the February 17th meeting of the Central Section are actually scheduled for the 
meeting on Thursday, March 16. The Section meets at 8:00 P.M. in the Western 
Society of Engineers' Auditorium, 84 East Randolph' St., Chicago. The first 
paper is "A New Amplifier Design" by Mr. Frank Mclntosh. The second paper, 
in two parts, describes the new DuPont Type 275 Color Release Positive Film. 
"Structure and Properties" will be described by Dr. A. B. Jennings, while "Print- 
ing and Processing" will be covered by Dr. J. P. Weiss. 

Pacific Coast 

On Tuesday, March 28, members of the Pacific Coast Section will be guests of 
the Eastman Kodak Company. They have been invited to attend the first 
'open house' tour of the Kodachrome Cine-Processing Laboratory, 1017 No. Las 
Palmas Ave., Hollywood, Calif. 


New Products 

Further information concerning the material described below can be 
obtained by writing direct to the manufacturers. As in the case of 
technical papers, publications of these news items does not constitute 
endorsement of the manufacturer's statements nor of his products. 

The Eastman 16-Mm Projector, Model 25, just announced by Eastman Kodak 
Company, has been developed to fill the need for a projector of professional caliber. 

Engineered with performance and long operating life as the major design objec- 
tives, this projector introduced a number of outstanding features. An 8-frame 
intermittent sprocket is driven by a synchronous motor interlocked but mechani- 
cally independent of a second motor that drives the film transport sprockets and 
the shutter. A geneva star gives intermittent film motion but is unconventional 
in that it is driven through a unique two-stage acceleration movement, which in- 
creases the ratio of sprocket speed to motor speed at the moment of film pull- 
down. This gives extremely high acceleration of the intermittent sprocket dur- 
ing the work period of its operating cycle. Shutter efficiency is therefore high. 


With two interruptions per frame, the shutter has a transmission of 59%. Indi- 
vidual motors are used to drive the film take-up and rewind reels. 

Either carbon arc or 1,000-watt tungsten lamps may be used. When tungsten 
lamps, which burn base up, are used, a dual lamp turret allows a stand-by lamp to 
be moved into position in the event of a burn-out. 

The sound scanning drum is driven by the film while a magnetically damped 
stabilizer holds the flutter content to 0.2% rms. 

The projector base houses an illuminated control panel and an Altec Lansing 
amplifier, with tone and volume controls as well as a switch for phonograph or 
microphone input. 

The projector is available, equipped for 115-volt, 60-cycle, a-c operation, and 
with either the Altec-Lansing Model 800 or the Model 604-B speaker. 

"Permanent and lengthy 'on-the-spot' 
sound recordings" is the lead of the 
release accompanying this photo from 
the Miles Reproducer Co., Inc., 812- 
814 Broadway, New York 3, N.Y. 
The equipment is described in part as 
follows: The "W T alkie-Recordair 
weighs only 8 Ib and is enclosed in an 
inconspicuous carrying case, measuring 
only 4 X 8 X 10 in., which conceals 
the identity of a recorder. It is de- 
signed to make permanent and continu- 
ous recordings, with a concealed micro- 
phone, of lengthy conversation of 
near-by and distant voices, while stand- 
ing, walking or riding on trains, autos 
or planes. 

Recording is reported to be noise- 
less, "thus no one is aware that con- 
versation is being recorded," and the 
operator is said to need to give only the 
attention required to throw a silent, 
hidden and external switch to "On" 
position. It is powered by small 
flashlight cells and by a miniature 
"B" battery, both of standard type. 

The manufacturer suggests that, in 
addition to some other, perhaps more 

obvious, uses "Walkie-Recordall" may be a boon in bolstering salesmen's 
potentialities: (1) through the aid of recorded expert sales talks to be played 
back to customers and (2) through reviewing and evaluating an entire day's 
actual conversation between a salesman and his customers. 

SMPTE Officers and Committees: The names of Society Officers 
and of Committee Chairmen and Members are published annually in 
the April issue of the JOURNAL. Changes and corrections to these 
listings are published in the September JOURNAL. 


Employment Service 


Project Engineer: Mechanical engi- 
neering graduate experienced in de- 
signing from specifications; optical 
instruments, precision cameras, me- 
chanical servo, and gear or 3-bar com- 
puters, analytical work in stress and 
vibration. R. A. Barbera, 663 Oving- 
ton Ave., Brooklyn 9, N.Y. 

TV and Motion Picture Engineer: 3 
yr experience in motion picture engi- 
neering and research at Philips Physi- 
cal Laboratories, Eindhoven; 6 yr as 
TV-Director, same firm; 3 yr as Di- 
rector of Decca plant in Belgium. De- 
sires assignment in any part U.S.A. 
Highest qualifications and references 
U.S. firms. Write Fernand Beguin, 
c/o Mr. Marc Albanese, 416 Madison 
Ave., New York 17. 

In technical phase: Motion picture or 
still photography. 4 yr experience in 
research, development, and testing, 
both color and b & w films. Gradu- 
ating from M.I.T. June, 1950. Mem- 
ber, SMPTE. W. A. Farmer, 141 
Grand Ave., Rochester 9, N.Y. 

Cameraman-Director: Currently em- 
ployed by internationally known pro- 
ducer, desires greater production oppor- 
tunities. Fully experienced 35- and 
16-mm, color, b & w; working knowl- 
edge editing, sound, and laboratory 
problems; administrative experience. 
Top references and record of experience 
available. Write P.O. Box 5402, Chi- 

Cameraman: Trained with practical 
experience in 16-mm and 35-mm equip- 
ment & technique with prominently 
successful men in the industry. Thor- 
oughly familiar with B & H Standard, 
Mitchell, Eyemo, & Filmo cameras, 
Moviolas, etc. Thorough knowledge 
& experience script-to-screen produc- 
tion technique: directing, editing, 
photography, film evaluation, produc- 
tion, treatments, shooting-scripts, small 
budgets, documentary & theatrical 
production. Go anywhere. Age 33. 
Top industry & character references 
furnished confidentially. Anxious for 
position where ability, sincere interest 
and creativeness offer opportunity. 
Active Member of SMPTE. Write 
Milton L. Kruger, R.F.D. 1, Ridge- 
wood, N.J. 


By action of the Board of Governors, October 4, 1931, this Honor Roll was 
established for the purpose of perpetuating the names of distinguished pioneers 
who are now deceased. 

Louis Aime Augustin Le Prince 

William Friese-Greene 

Thomas Alva Edison 

George Eastman 

Frederic Eugene Ives 

Jean Acme Le Roy 

C. Francis Jenkins 

Eugene Augustin Lauste 

William Kennedy Laurie Dickson 

Edwin Stanton Porter 
Herman A. DeVry 

Robert W. Paul 
Frank H. Richardson 

Leon Gaumont 

Theodore W. Case 

Edward B. Craft 

Samuel L. Warner 

Louis Lumiere 

Thomas Armat 


Lee de Forest 

A. S. Howell 

Journal of the Society of 

Motion Picture and Television Engineers 


The Eidophor Method for Theater Television E. LABIN 393 

Standard Television Switching Equipment RUDY BRETZ 407 

Color Temperature: Its Use in Color Photography 0. E. MILLER 435 

An Experimental 35-Mm Multilayer Stripping Negative Film 


Printing Equipment for Ansco Color Film F. P. HERRNFELD 454 

16-Mm Film Color Compensation 0. K. KENDALL 464 

Illuminating System and Light Control for 16-Mm Continuous Optical 


New Brenkert Projection System for Drive-in Theaters 

C. N. BATSEL and H. J. BENHAM 483 

Note on Metol Analysis in Photographic Developers . MARTIN IDELSON 492 


16-Mm Camera Apertures, Z22.7; 16-Mm Projector Apertures, Z22.8; 
8-Mm Camera Apertures, Z22.19; 8-Mm Projector Apertures, Z22.20; 
Frames for Theater Screens, Z22.78; 16-Mm Sound Projector Test Film, 

Society Announcements 508 

Engineering Committees 509 

Current Literature 510 

BOOK REVIEW: Noise and Sound Transmission A Symposium 

Reviewed by Vern O. Knudsen 511 

New Products 512 

Employment Service 513 


Administrative Committees 515 

Engineering Committees 518 

Representatives to Other Organizations 522 


Chairman Editor Chairman 

Board of Editors Papers Committee 

Subscription to nonmembers, $12.50 per annum; to members, $6.25 per annum, included in 
their annual membership dues; single copies, $1.50. Order from the Society's General Office. 
A discount of ten per cent is allowed to accredited agencies on orders for subscriptions and 
single copies. Published monthly at Easton, Pa., by the Society of Motion Picture Engineers, 
Inc. Publication Office, 20th & Northampton Sts., Easton, Pa. General and Editorial Office, 
342 Madison Ave., New York 17, N.Y. Entered as second-class matter January 15, 1930, 
at the Post Office at Easton, Pa., under the Act of March 3, 1879. 

Copyright, 1950, by the Society of Motion Picture Engineers, Inc. Permission to republish 
material from the JOURNAL must be obtained in writing from the General Office of the Society. 
Copyright under International Copyright Convention and Pan-American Convention. The 
Society is not responsible for statements of authors or contributors. 

Society of 

Motion Picture and Television Engineers 

342 MADISON AVENUE, NEW YORK 17, N.Y., TEL. Mu 2-2185 

Earl I. Sponable 
460 W. 54 St. 
New York, 19, N.Y. 


Peter Mole 

941 N. Sycamore Ave. 

Hollywood 38, Calif. 



Loren L. Ryder 
5451 Marathon St. 
Hollywood 38, Calif. 

Clyde R. Keith 
120 Broadway 
New York 5, N.Y. 

Robert M. Corbin 
343 State St. 
Rochester 4, N.Y. 

William C. Kunzmann 
Box 6087 
Cleveland 1, Ohio 

Fred T. Bowditch 
Box 6087 
Cleveland 1, Ohio 


Frank E. Cahill, Jr. 
321 W. 44 St. 
New York 18, N.Y. 

Ralph B. Austrian 
25 W. 54 St. 
New York 19, N.Y. 


Herbert Barnett 
Manville Lane 
Pleasantville, N.Y. 

Kenneth F. Morgan 
6601 Romaine St. 
Los Angeles 38, Calif. 

Norwood L. Simmons 
6706 Santa Monica Blvd. 
Hollywood 38, Calif. 


George W. Colburn 
164 N. Wacker Dr. 
Chicago 6, 111. 

Charles R. Daily 
113 N. Laurel Ave. 
Los Angeles 36, Calif. 

John P. Livadary 
4034 Cromwell Ave. 
Los Angeles, Calif. 

William B. Lodge 
485 Madison Ave. 
New York 22, N.Y. 

Edward Schmidt 

304 E. 44 St. 

New York 17, N.Y. 


Lorin D. Grignon 
1427 Warnall Ave. 
Los Angeles 24, Calif. 

Paul J. Larsen 
6906 S. Bradley Blvd. 
Bradley Hills 
Bethesda, Md. 

William H. Rivers 
342 Madison Ave. 
New York 17, N.Y. 

Edward S. Seeley 
161 Sixth Ave. 
New York 13, N.Y. 

R. T. Van Niman 
4501 Washington St. 
Chicago 24, 111. 

The Eidophor Method 
For Theater Television 



Summary The paper considers briefly some system aspects of theater 
television. The most important single technical problem of theater television 
remains the method used for large-screen projection. A new system, devel- 
oped hi Switzerland, using a special electromechanical accumulation process 
is described briefly. It gives projection on full-size screens with ordinary 
lighting. The system is still in the stage of early development, but has al- 
ready proven quite promising. 

/CONCEPTIONS of how theater television will be used are not yet 
\^l clearly settled. This is partially due to the limitations of the 
available technical tools, and partially to the difficulties of under- 
standing how theater television will fit in with the normal movie 

Figure 1 represents a possible network for theater television. It is 
set up of urban, or local, networks, interconnected through long dis- 
tance communication links. In each local network there would be 
one or more centers. The pictures originating at special studios in the 
city (stadiums, theaters, nightclubs, or perhaps as received from 
broadcast studios) are received in the local center after transmission 
through local telephone facilities, or, more likely, through short range 
microwave links. In this television distribution center, the pictures 
may be recorded on films, or may be rerouted immediately to the 
various theaters in the city and may also be sent through long distance 
networks to another city. In other words, the center operates like a 
telephone central office, receiving the messages from various points 
and retransmitting them to the customers. In a city where numerous 
theaters are grouped company-wise, it may be imagined that each 
company would have one central office of this type. In other cities, 
one central office may operate as a sort of limited common carrier for 
various theater owners. The long distance intercity links may, in 
turn, be operated directly by theater owners or may be facilities 
rented from a common carrier company. 

The technical tools for setting up a network of the type described 

PRESENTED: October 14, 1949, at the SMPE Convention in Hollywood. 





in Fig. 1 can be grouped in three main categories: transmission, 
pickup or camera, and projector. The transmission problems can be 
solved with known methods and it is now possible to equip, with 
microwave links or coaxial cables, both the local and the long distance 
networks. While continuous progress is still to be expected in the 
design of improved cameras, it is well known that cameras are already 
available with excellent sensitivity. The situation concerning pro- 
jectors is not quite as clear. Two systems have been put in experi- 
mental operation: the intermediate-film method and the direct- 
projection method using cathode-ray tubes. Another direct-projec- 
tion method has been developed at the Polytechnical Institute of 
Zurich, Switzerland, and has been demonstrated successfully in ex- 
perimental form. 

Figure 1. 

The main purpose of this paper is to describe the Swiss system. I 
want to make it quite clear that I have no direct connection with the 
development of this system. That was done entirely at the Poly- 
technical Institute of Zurich under the direction of Professor Fischer, 
at the beginning of the project. 1 Since Professor Fischer's death in 
1948, the project has been continued under the very able direction of 
Professor Baumann. I am reporting on this system at the request of 
Mr. D. E. Hyndman, Chairman of the SMPTE Theater Television 
Committee. The only justification for me to report about this sys- 
tem is that I saw it in operation last year in Zurich and I have taken a 
great interest in these developments. This paper is developed from 


material prepared by Dr. H. Thiemann, 2 who is Professor Baumann's 
assistant and is in charge of this development in Zurich. It is through 
Dr. Thiemann's courtesy that I have been authorized to illustrate this 

If the network conception, briefly described above, is correct, both 
methods of theater projection, by the intermediate film or through 
direct projection, are necessary. At the central station office the 
intermediate-film process is obviously an indispensable element. 
A permanent record is required for retransmission to the various sub- 
scribers at convenient times. On the contrary, at the subscriber's 
theater there is no need for recording and therefore a direct-projection 
system is probably the most desirable solution. So far, the only 
practical solution for direct projection is to use a powerful cathode- 
ray tube. 

How much power do we have to concentrate in the electron beam in 
order to illuminate the theater screen properly? For a screen of 18 X 
24 ft with a luminance of 10 foot-lamberts assuming reflection con- 
forming to Lambert's law with a coefficient of 0.7, the amount of light 
flux required is 6,000 lumens on the screen. If this power is pro- 
jected through a refractive optical system which would have an over- 
all efficiency of 3%, the source would have to supply 200,000 lumens. 
With reflection-projection systems of the Schmidt type, one can in- 
crease the over-all efficiency to 30% and therefore the light to be sup- 
plied by the source would have to be 20,000 lumens. If the trans- 
formation from electron beams into light could be done without any 
loss and if that transformation could take place at the wavelengths for 
which the eye is most sensitive, 620 lumens would correspond to one 
watt in the electron beam. The power required in the electron beam 
would therefore be 30 watts for a transformation efficiency of 100%. 
The transformation efficiency is actually not more than 1% or 2%, 
being the product of the spectral efficiency and of the energetic effi- 
ciency. The spectral efficiency is of the order of 10% because the 
light produced by the electron beam is not at the wavelength for which 
the eye is most sensitive. The energetic efficiency of the transfor- 
mation from electrons to light quanta is also of the order of 10%, re- 
sulting in a final efficiency of 1%, or perhaps 2%. 

Finally, the power required in the electron beam is of the order of 
3 kw. Present-day cathode-ray tubes do not handle such large 
amounts of power in the beam. One is, therefore, obliged to cut cor- 
ners to accept a lower screen brightness, or to try to increase the re- 

396 E. LABIN April 

flection coefficient of the screen by using directive screens, or to use 
smaller screens. 

In all cases, for large screens one has been led to the Schmidt pro- 
jection system because of its great effective aperture but this system 
has some limitations, the most disturbing one being that the distance 
between the projection optics and screen cannot be increased without 
requiring an extreme mechanical accuracy in the projection system. 
Present-day practice does not allow a throw as great as the one nor- 
mally used in theaters and therefore the projector cannot be installed 
in the normal projection booths. Improvements in cathode-ray tube 
projectors can be expected, but the figures quoted above show that, at 
best, one could hope to catch up with present-day practices in movie 
theaters; but the tendency is for more light and larger screens. The 
cathode-ray projection scheme has enough limitations to justify an 
attempt to look into the possibility of developing another competi- 
tive method. The fundamental philosophical objection one could 
raise against the cathode-ray system is that light is generated by the 
electron beam in the cathode-ray tube and is also controlled in the 
cathode-ray tube itself. Many proposals have been made in the 
past, based on the general idea that the light energy will be supplied 
from a source, such as a projection lamp, and that the intensity of the 
light which reaches the screen will be controlled by some independent 
modulation device. As far as we know, none of these proposals has 
been actually made to work satisfactorily, except the system using the 
eidophor control of Professor Fischer. 

The principle of the system can best be understood with reference 
to Fig. 2. The light from an arc is projected on the eidophor and the 
surface of the eidophor itself is projected on the screen. Between the 
light source and the eidophor there is a slit system (or Schlieren 
optics), and a second one is located between the eidophor and the 
screen. The eidophor, which represents the control element, is a thin 
layer of viscous liquid deposited on a very thin metallic electrode 
which is transparent to the light beam. The eidophor is mechanically 
deformed by electrostatic forces resulting from charges which are de- 
posited on the surface by an electron beam hitting the eidophor at a 
certain angle. The charge produces electrostatic stresses, and the 
corresponding deformation of the surface, together with the two series 
of slits, makes it possible to control the light flux. 

The theory of the system is based on the assumption that actual 
diffraction with phase coherence takes place when the light crosses the 




eidophor. Figure 3 represents a simplified hypothetical case where 
one slit only is used instead of a number of bars. In the absence of 
corrugation on the eidophor surface the lens forms a single image of 
the lower slit in the plane of the "picture slits." When regular cor- 
rugations are present additional secondary images are formed by dif- 
fraction, as shown in Fig. 3, in a manner similar to the secondary 
images obtained from familiar diffraction gratings. The secondary 
images are displaced by an angle B which depends only upon the wave- 
length of the light used and the period of the corrugation on the sur- 
face of the eidophor. 

Fig. 2. Large screen 
television projector. 


Fig. 3. Diffraction through control grids. 

The distribution of light intensity in the secondary images deter- 
mines the efficiency of the light control which can be achieved in this 
manner. This light distribution can be calculated and it can be shown 
that the zero order or original image can be extinguished completely 
while most of the light flux is shifted into the secondary images. Nor- 
mally, the secondary images of the first order only are used. The 
study of the single slit arrangement of Fig. 3 for various forms of de- 
formations of the eidophor surface (sinusoidal, triangular, rectangular, 
etc.) leads to a certain number of light control curves. A similar 




study has been conducted for the extreme opposite theoretical case 
where an infinite number of slits are used. 

From the theoretical curves thus obtained a choice of the optimum 
number of slits and of their geometric configurations has been made 
corresponding to the type of deformation occurring in a television 
picture. For a complete analysis of this phenomenon, which would 
go far beyond the limits of this presentation, the reader is referred to 
the original paper of Dr. Thiemann. 

Figure 4 shows one possible arrangement. When the surface of the 
eidophor is flat, the slits are so arranged that no light is transmitted; 
this, therefore, corresponds to a black picture. When the surface of 
the eidophor is deformed with a sinusoidal undulation, the total 
amount of light transmitted corresponds to a white picture. A series 
of black-and-white lines in the picture would correspond to a defor- 




Fig. 4. Deformations 
on the eidophor. 

mation of the surface of the eidophor which would be a flat portion 
followed by an undulated portion. 

The period of the undulation required for a white picture has, of 
course, to be smaller than the duration of one picture element. In 
order to generate the deformation required on the eidophor, the elec- 
tron striking the eidophor has to be modulated by the video signal. 
One possible method would be to modulate the beam intensity with a 
high-frequency carrier. For a white picture, the carrier would have 
the full amplitude and for a black picture the carrier would be en- 
tirely suppressed. This method of modulation is not the one that has 
been finally chosen because, the modulation curve not being fully 
linear, any intensity modulation is connected with some rectification. 
This rectification gives rise to charges on the surface of the eidophor. 
The charges, in turn, generate a constant deformation of the liquid 
surface and, therefore, an image once registered remains on the sur- 




face in more or less a rudimentary form for a considerable time. In 
order to eliminate this cause of disturbance, a different kind of modu- 
lation has been chosen. The intensity of the beam remains constant 
but the scanning speed in the direction of the line is varied to conform 
to the video signal. 

The video signal is still superimposed on a carrier which will act on 
the deflection of the electron beam through additional deflection 
plates. (The deflective voltage required is 1 volt.) In order to gener- 
ate the desired deformations of the eidophor within our picture ele- 
ment the spot size of the electron beam shall be no more than one- 
quarter of the desired local period of the deformation which in turn 
shall be close to one-half the size of the picture element. This leads 

Fig. 5. Light control curve 
of the Schlieren optical system. 

' (to* /M 

20 30 40 

to very fine cathode-ray beams (the spot size is .001 in. in the direction 
of the scanning lines) and has represented one of the difficult problems 
of the system. 

Figure 5 shows the modulation curve which can be actually 
achieved. The variation of light intensities is shown in a logarithmic 
scale versus the video input signal. In the curve shown, the effect of 
stray light has also been taken into consideration. The great impor- 
tance of this stray light which prevents reaching complete black is 
clearly shown. Calculation of the efficiency on the complete control 
system shows a value to 40% for full modulation. Taking into ac- 
count the fact that 50% of the light is absorbed by the slit system, 
the net light efficiency is approximately 20%. In other words, for a 



light flux of 6,000 lumens on the screen, one would need an arc lamp 
of 30,000 lumens. For good light control the requirements of mini- 
mum stray light is not easily met. Many secondary causes may add 
to the ultimate amount of stray light, but the most important of all 
sources of remaining background light is reflection on glass surfaces. 
It is absolutely necessary to make all the glass surfaces within the 
Schlieren optics free from reflection. The laboratory equipment 
built at the Zurich Polytechnical Institute has met these requirements 
very well. Actually, one of the major qualities of the pictures demon- 
strated in Zurich is the excellent contrast which can be achieved com- 
bined with a smooth gradation curve. It seems to be rather superior 
to the pictures which can be produced with most cathode-ray tubes. 

Figure 6 shows the block diagram of the large screen projector. 
The figure is self explanatory. The design and construction of the 
various electronic parts shown in the block diagram have not always 
been easy; but the problems they raise are of conventional nature and 
it does not seem necessary to present detailed comments concerning 
this part of the system. The preparation of a suitable eidophor liq- 
uid presents a much more difficult and original problem. The liquid 
has to meet very stringent specifications. Since it operates in high 
vacuum, its vapor pressure has to be very low, if possible appreciably 
lower than 10~ 5 millimeters. The conductivity, viscosity, dielectric 
constant and capillarity have to be inter-related by a specific relation 
which can be found from a theoretical study of the modulation con- 
ditions. The color of the eidophor must not have a disturbing effect. 
Finally, the eidophor has to withstand the bombardment of the cath- 
ode beam without being destroyed. After considerable experiment- 
ing, a suitable liquid has been found which seems to be reliable and 
which does not seem to be altered after long periods of operation. 
This is basically a mineral oil with suitable additions to bring the 
conductivity up to the desired value. 

The arrangement as used thus far is shown in Fig. 7. The eidophor 
liquid is spread on a glass plate on which there is applied an electrode 
in the form of a transparent metallization. The glass plate moves 
very slowly (one turn in many minutes). After the liquid has been 
used for several picture frames, it is brought into contact with a cooled 
metal plate. After it leaves the cooling plate, it is smoothed out again 
by means of a straight edge or a rake into an optically perfect surface. 

The movement of the eidophor is intended to permit cooling of the 
eidophor and to avoid disturbing border phenomena produced by 



Fig. 7. Eidophor carrier 
with rake and cooling arrangement. 


Fig. 8. Detail of the completed large screen projector: cathode-ray tube on top 
and center; eidophor plate holder below; holder for output slit system in back of 
cathode-ray tube. 



electrostatic charges which would be accumulated on the eidophor if 
it were continuously submitted to the electron bombardment. 

The various requirements briefly mentioned above for operation 
of the eidophor and the accuracy of the design for the optical system 
have led to a first experimental model which is quite complicated and 
is in no way intended for commercial applications. 

Figures 8, 9 and 10 are pictures of some of the elements of the com- 
plete machine and show how complex and large a structure it is. 
The plate holder shown in Fig. 9 is approximately 6 ft in diameter. 
The whole machine occupies two floors, the arc lamp being at a floor 
below the projection room. The complexities of the present installa- 
tions should not prejudice the possible future applications. Professor 
Baumann has worked out a project for a new model and has authorized 
me to present the expected over-all dimensions of this new machine, 
shown approximately in Fig. 11 as: height, 5 ft; length, 5J^ ft; 
width, 2>^ ft; and weight 1800 Ib. 

The great saving in size and volume as compared to the present sys- 
tem is due to a new conception of an optical system as it appears from 
Fig. 12. The light, instead of traversing the eidophor once as in the 
present system, will be reflected by the parabolic mirror and will re- 
turn on the same slit system. Many advantages are expected from 
this new arrangement, such as continuous cooling by the parabolic 
mirror, very slow motion of the mirror (one revolution in one hour), 
inexpensive optical system, very high contrast ratio (1 to 300), and 
convenient aperture of the final light beam (1/7.4). With an arc of 
approximately 70,000 candles per square centimeter, Prof. Baumann 
expects to get a maximum light flux of approximately 7,000 lumens on 
the screen. Like the first laboratory model, the whole equipment 
will be evacuated continuously by means of a rotary pump. 

From these very brief descriptions it can be seen that the eidophor- 
projection system, while still in the experimental stage, is progressing 
toward a more practical solution. The new project on which the ex- 
perts of the Polytechnical Institute of Zurich are working is quite 
promising and would represent an equipment which in complexity 
and size is not very different from the latest projection systems using 
cathode-ray tubes. It should be noted that the Swiss system is in- 
tended for installation in the normal projection booth of the movie 
theater and from that point of view it has a considerable practical 
advantage over the cathode-ray tube systems using Schmidt optics. 

To conclude, I would like to remind you once more that the remark- 

Fig. 9. Detail of completed large screen projector: eidophor plate 
holder at left; pumps' connection at right; cooling liquid connection 
at extreme left. 

Fig. 10. Detail of completed large screen projector: arc lamp; 
optical system before reaching eidophor (eidophor holder is on floor 




Fig. 11. Projected plant. 

Fig. 12. Projected plant diagrammatic structure, 
with radar-purpose mercury high-pressure lamps. 

406 E. LABIN 

able development which I have tried to summarize is being under- 
taken entirely at the Poly technical Institute of Zurich by Professor 
Baumann, Dr. Thiemann and their colleagues. At the request of the 
SMPTE Theater Television Committee, I have tried to give a short 
description of this system. I do not know if it will ever be a com- 
mercial competitor for the cathode-ray tube projection system, but it 
would be very surprising if such a remarkable new tool would not find 
some useful applications. I want to thank Mr. D. E. Hyndman of 
the Theater Television Committee for the opportunity he gave me to 
talk at the SMPE Convention and I also wish to express my gratitude 
to Dr. Thiemann who supplied the material for this paper. 


(1) F. Fischer and H. Thiemann, "Theoretische Betrachtungen uber ein neues 
Verfahren der Fernsehgrossprojektion," Schweizer Archiv, vol. 7, January, Febru- 
ary, November and December, 1941; vol. 8, May, June and July, 1942. 

(2) H. Thiemann, "Fernsehgrossprojektion nach dem Eidophorverfahren," 
Bulletin De L' Association Suisse des Electriciens, p. 585, No. 17, August 20, 1949; 


MR. W. W. LOZIER: Approximately what was the picture size on the eidophor? 

DR. LABIN: Approximately twice the normal motion picture size. 

MR. LOZIER: Can you give any lumen values of what you have obtained on, say, 
a white screen? 

DR. LABIN: Yes, we have obtained 1500 lumens on a screen which was ten 
meters square. I am quoting from memory of the papers. 

MR. LOZIER: Do you know approximately what speed projection lens was used, 
whether there are any limitations there? 

DR. LABIN: No, I don't think there are limitations. They impose themselves 
to work with a normal angle of projection used in theaters. I mean they consider 
it as a must to install their machine in the normal projection rooms. It would be 
considered as a must, in view of the size of the machine. They have no other 
limitations that I see. I do not have the final efficiency in terms which would even 
indirectly answer your question. 

MR. LOZIER: That is what I was thinking. You would have a focal length, I 
think, of roughly twice what we use in theaters now. If you have a picture twice 
as big to start with and speeds in those focal lengths might run to //3 or //2, will 
the rest of the system fill such a speed? Can it be designed to do that? 

DR. LABIN: Oh yes, there is no limitation to that. If you question the actual 
optical efficiency of the system, it is certainly not as good as normal projection. 

Standard Television 
Switching Equipment 




Summary There is a considerable difference in design among the standard 
switching systems put out by the three major television manufacturers. 
General engineering departments of the major networks and many of the 
independent stations have designed and built their own systems. All 
studio switching systems should permit the operator to fade to black, lap- 
dissolve and superimpose. Not all will permit more complicated effects such 
as cutting to a superimposure, or cutting away from a superimposure to a shot 
on another camera. Not all will permit the operator to preview a super- 
imposure before he makes it. The following article lists the requirements 
of television switching systems from the operating point of view and de- 
scribes the -operation of the standard models which are in use in television 
stations today. 

Positioning of the Switching System 

Most control rooms are laid out in a two-tier arrangement with 
video engineers and camera control units on the front and lower tier. 
The second level provides a table for the director and usually for an as- 
sistant as well. The location of .the audio engineer varies between 
one level and the other, or he may be placed at the side of the control 
room, not in either tier. The bank of buttons and other controls, 
known as the switching system, is also found sometimes below and 
sometimes above. Figure 1 shows four different methods of control 
room layout. The technical director (T.D.) may sit at the same 
console as the video engineers (a), or he may sit beside the director 
(b, c, d) . NBC (d) likes to place the video men and camera monitors 
off to the side and leave the T.D. and the director alone in front of 
the control room window. Only a master monitor and one preview 

A CONTRIBUTION: Submitted January 10, 1950. This is part of a forthcoming 
book and is published by permission of McGraw-Hill Book Co., Inc. Critical dis- 
cussion of this material is particularly invited, either in the form of Letters to the 
Editor or by communication directly to the author at Croton-on-Hudson, N.Y. 






[ASST.D1R.] f DIR. I [rtSSrDH.| 

() ( b ) 


Fig. 1. Four methods of control room layout. 

Fig. 2. RCA Switching Panel Type TS-1A. 

monitor are used by the T.D. and the director under the NBC sys- 

The manufacturers of switching equipment usually build their 
switching systems into consoles, in some cases combined with a mas- 
ter monitor and in other cases with a preview monitor as well. The 
RCA studio switching system (Fig. 12), which is combined with a 


master monitor in this way, can be removed from the monitor if de- 
sired. Stations sometimes mount the switch panel in the production 
table on the second tier, leaving the master monitor down below. 
Most of the specially built switching systems are installed in this 

In the remote truck that WBKB has designed and built (one of the 
best designed trucks in the industry), there are two sets of buttons: 
one below at the T.D.'s desk, and one on the upper desk for the pro- 
ducer. It is possible for the T.D. to punch a "remote control" but- 
ton and throw the switching operation entirely to the director of the 
program, who then operates from his own set of buttons. 

Incoming Picture Signals 

A studio switching system must be designed to handle more in- 
coming picture signals than merely those emanating from the cameras 
in the studio. Projection equipment is often used for titles or film 
portions of studio shows, and the film channels must also be controlled 
from the studio switching system. In small stations, the studio con- 
trol room functions also as the film control room, sometimes actually 
containing the projection equipment in a back corner. In such cases, 
the film channel or channels normally feed through the studio switch- 
ing system. 

By-Passing the Studio Switching System 

When a film or test pattern slide is on the air and feeding through 
the studio switching system, the studio cannot be used for camera 
rehearsal. Any switching would disturb the program on the air. 
For this reason, there is always some means provided for by-passing 
the studio switching system and feeding directly from film camera to 
transmitter. The same by-passing is provided for other incoming 
signals, such as network programs from AT&T (if the station is on 
the coaxial cable), or a picture from a remote pickup. A series of 
master switching buttons is installed for this by-passing purpose, in 
the circuits between the studio switching system and the transmitter. 
Sound, of course, must be handled identically, but audio is a parallel 
system and not the concern of this chapter. Figure 3 shows the 
master switching panel for by-passing the studio switching system. 
Buttons shown in black are on. The film camera is feeding through 
the by-pass circuit directly to the transmitter, while the switching 
system is feeding only the client's booth and monitors. 




Preview Switching 

It is always a great risk to put anything on the air without being 
able to watch it right up to the moment of the switch. Any picture 
which does not appear on a camera-control monitor in the control 
room must be previewed somewhere before being switched onto the 
air. Film channels which are monitored and controlled elsewhere 
will have to be previewed in this way. When remote pickups are to 
be integrated with studio presentations, they also must be previewed. 



Fig. 3. Method of by-passing studio switching system. 






Fig. 4. Mixer type of switching system. 

This is particularly true on occasions when live or film commercials 
from the studio must be inserted into ball games or other sports 
events. At these times, the ball game will sometimes be fed through 
the studio switching system on its way to the transmitter, so that 
the studio-originated portions can be cut in. Any good switching 
system must provide a method of selecting the channels that are to 
be seen on the preview monitor. 


Some switching systems, which have a master monitor built into 
the same unit, make it possible to use the master monitor for preview- 
ing these incoming lines. This is a poor practice, however, since it 
sacrifices the master monitor during the time another channel is 
being previewed. It results in a type of previewing which amounts to 
only a quick glance at the next picture, and it certainly is not desir- 
able for best programming results. 


A switching system must also be designed to fade channels or cross- 
fade them so that dissolves, fades and superimposures can be ac- 
complished. There are three common ways to design a switching 
system to do these things. 

L The Mixer Type. The first system provides each channel with a 
separate gain control, and runs them all together into a mixer, just 
as the outputs of many microphones are mixed in an audio console. 
This kind of switching system may have separate switches, as well 
as separate fading controls. The old Mt. Lee studio of KTSL had 
such a switching system, built according to the studio's own design. 
It had seven positions for fading, but no switches. Instantaneous 
cuts had to be approximated by, very quick dissolves. The Dumont 
"mixer," which is a switching system of this general type, provides for 
only four channels, but has switches and a number of additional fea- 
tures and refinements (Fig. 4). 

2. The Dual-Fading-Bus Type. The second type of switching sys- 
tem provides two basic master channels, which feed through two 
fader controls. They are usually termed "channel" or "fading bus" 
A and B, and all of the video channels which feed into the switching 
system can be punched up on either one (Fig. 5). 

Only one of these two channels is used when straight switching is 
desired ; the fader control for the channel being used is left open and 
the fader control for the other is closed. If you have been using 
Channel A, for example, and wish to make a dissolve, first punch up 
the camera you will dissolve to on Channel B, then simultaneously 
fade out Channel A and fade in Channel B. There will be a further 
discussion of this later in this chapter under "Specific Equipment." 
The RCA studio switching system is of this design, as well as the 
General Electric switching system which is built into a program con- 




3. The Three-Bus Type. The third system has three master chan- 
nels, two of which are for fading and dissolving, as just described, 
while the third is used for straight switching. The straight switching 
bus has an extra button marked ' 'Effects," through which the com- 
bined output of the two fading buses will feed whenever dissolves or 
superimposures are desired. This has an advantage over the second 

Fig. 5. Dual-fading-bus type of switching system design. 





Fig. 6. Three-bus type of switching system design. 

type of switching system in that it permits switching to a super- 
imposure or switching away from a superimposure to a single camera 
(Fig. 6). 

This type of switching system was developed by NBC and has 
been installed in all the NBC-owned stations. The RCA switching 
system, which is built into the Program Director's Console, is of this 
design. CBS has made some very useful additions to this equipment. 



The RCA Field Switching System 

The RCA Field Switching System, first on the market in 1946, is 
very frequently seen, not only as remote equipment, but in studio use 
as well. This equipment originally had no provision for making 
dissolves, and several ingenious devices have been developed to adapt 
it to this purpose. RCA now offers a special auxiliary unit which 
is equipped with dual fading buses similar to the RCA studio switch- 
ing system (Fig. 7a). 

The control face of this switching system is divided into three 
horizontal areas. At the top is a bank of 13 switches which control 
a very complicated and flexible intercommunication system, de- 
scribed in detail in the chapter on intercommunication. Beneath 
the cover in the middle portion is an intercom jack panel, containing 
six plug-in points for the engineers' and director's headphones. 

The lower third only is concerned with switching of picture signals. 
There are two rows of buttons with associated tally lights to show 
which channel is on the air. The top row is for monitor switching 
and the bottom row controls the outgoing picture line. The monitor- 
switching feature, as explained above, makes it possible to use the 
master monitor as a preview monitor when working with limited 
equipment. The master monitor is, of course, a separate unit. 

Six incoming signals are provided for in this switching system. 
Four of these are intended for camera channels and two (auxiliary 5 
and 6) for incoming channels, usually from a remote source. 

Monitor switching allows the master monitor to be used for several 
purposes in addition to its regular use as a master monitor on the 
outgoing picture line. Either of the incoming pictures lines, 5 or 6, 
can be previewed by punching the appropriate buttons. The engi- 
neer has occasional need for checking the picture at various points 
in the system for instance, at the input and output of the relay 
transmitter when this equipment is being used in field pickup; 
provision is made for these purposes also. In actual practice, on a 
remote pickup, monitor switching is used only in case of emergency. 
It would be extremely confusing to a director not to know if the 
picture being shown on the master monitor is on the air. 

The knob above the monitor switching buttons is concerned with 
the two auxiliary lines 5 and 6. When a picture is coming in from a 
remote source, say another studio or a field pickup, the equipment in 




Fig. 7a. Control panel of RCA Field Switching System, 
Type TS-30A. 

that location probably includes its own synchronizing generator, 
and the incoming signal is a composite signal ; that is, the synchro- 
nizing pulses are already combined with it. In the case of the camera 
channels, the signals coming in to the switching system are pure video, 
and the synchronizing pulses are added in the switching system just 


before the program goes out to the transmitter. When the incoming 
signals on lines 5 and 6 are complete with synchronizing signals, the 
switch knob is set to EXT (external sync). If, instead, picture 
signals from additional cameras or film pickup chains are fed through 
the auxiliary lines, these are not composite signals and synchronizing 
signals must be added. For this the switch is set to INT (internal 

Dissolves with the RCA Field Switching System. The TS-30A 
switching system provides for straight switching only, since when it 
was designed in 1945 no one anticipated the need for dissolves or 
superimposures in the field, and studio use of this equipment was not 

A great many small stations have installed this field equipment 
in their studios, however, because of its low cost. Naturally, the 
requirements of studio production make dissolving, at least, a neces- 
sity. In the field an even greater use has been found for the super- 
imposure than for the dissolve. During brief intermissions in a game, 
the director may want to show the sponsor's name or symbol on the 
screen as a visual complement to a short commercial, without at the 
same time giving the audience the feeling of losing contact with the 
field, where action may begin again at any moment. In such cases, 
a superimposure is very valuable. A very striking effect is often 
achieved by superimposing a close-up of a beer bottle, for instance, 
on the baseball field in such a way that it looks like a gigantic bottle 
actually resting on the field. Perspective must match to achieve 

It is possible, with a little co-ordination between video engineer 
and technical director, to use the equipment as it now stands to ef- 
fect a very credible dissolve. It has been found that two of the video 
switching buttons can be depressed at the same time, putting two 
signals on the air simultaneously. The result of this is, of course, a 
sudden superimposure when the second channel is added, an effect 
which is very rarely of any value. But if the second channel is added 
in a faded-down condition, and' then the gain brought up after it 
is on the line, a slow appearance of the superimposed shot is effected. 
If the first picture is faded down at the same time that the second 
picture is brought up, a dissolve is the result. The gain controls are 
not associated with the switching system, but are found on the camera 
control units. Consequently, a very exact co-ordination is necessary 
between technical director and video engineer to make this result 




possible. I have seen it done by one man in the studio at KTSP- 
TV, where only one engineer was available for all the video functions 
(Fig. 7b}. 
The routine, with the proper director's cues, is as follows : 

1. "Fade down Two." (Assuming that Camera One is on the air, 
the desired shot is framed up on Two and the cameraman is told to 
hold it, while the instruction to fade down is given.) 

2. "Add Two." (Camera Two is punched up on the switching 
system, but Button One must be held down, so that it won't auto- 
matically switch off since all buttons in each row are mechanically 
interlocked. Since Two is now faded down, there is no visual effect 
except a slight tonal change due to the added circuit.) 

3. "Dissolve." (The dissolve can now be effected by simultane- 
ously fading down the gain on One and fading up the gain on Two. 

Fig. 7b. Simple metjiod of making dissolves 
with RCA Field Switching System. 

To be properly co-ordinated, this must be done by one person, and of 
course, he must watch the master monitor during the process. It may 
be found, for instance, that it is best to fade in Two about halfway 
before starting to take One out, or a dark period will be noticeable in 
the middle of the dissolve.) 

4. "Drop One." (It is now necessary to take Camera One off the 
line, before any cuts are made, unless a dissolve back to One is ex- 
pected. To take out One, Button Two must be held down while one 
of the other buttons is depressed just half way, far enough to trip 
the automatic release which snaps out Button One. 

5. "Fade up One." (When the dissolve was effected, Camera One 
was faded down to black; before it can be used again, the gain must 
be turned up. This final step is frequently forgotten in learning this 
method of making dissolves.) 

A strange effect is noticeable on the camera monitors when two 
switching system buttons are held down. The picture from each 




camera feeds back through the line from the other camera and ap- 
pears on that camera's monitor as well. This is not noticeable if the 
above routine is carried out properly, since one camera is always 
faded down while two buttons are depressed. 




^ j-r"? LL, 


000000 / P 

O 0^ I r^T? 




Figs. 8, 9, and 10. Adaptations of RCA Field Switching 
System to permit dissolves. 

Special Adaptations of the Field Switching System. The method of 
effecting dissolves just described requires more co-ordination be-* 
tween various operators than most television stations like to rely on. 

418 RUDY BRETZ April 

A self-contained switching and dissolving unit is much simpler to 

One method of adapting the field switching system for making 
dissolves was devised by WENR-TV in Chicago (Fig. 8). That sta- 
tion mounted separate potentiometers (fading dials) for each of four 
channels in the space where the intercom plug board is located. 
This method changed the switching system to a mixer type of dissolv- 
ing system, similar to the Dumont mixer. The intercom facilities 
were, of course, sacrificed, but that was not serious, since most studios 
prefer to use their own separate intercom systems. 

A second method is to convert the equipment into a dual-fading- 
bus type of switching system; one bus is the regular set of program 
selector switches, the second bus is the set of monitor selector switches. 
A separate potentiometer can be connected to each of these fading 
buses and mounted on the face of the switching system, or two 
potentiometers can be mounted face to face on one shaft, so that 
they work in opposite directions and only one dial is necessary. 
With two potentiometers on a single shaft, however, all that can 
be accomplished is a dissolve, since one channel fades in as the other 
fades out. With separate control, fade-outs and superimposures 
can also be effected (Fig. 9). 

A third method is to feed only two channels into a separate dissolve 
control. This method may make use of a pair of potentiometers on 
the same shaft, or two separate dials, mounted in an auxiliary box. A 
certain number of studios are equipped with complicated versions 
of this "dissolve box," some containing separate dual-fading buses 
and camera-selector switches. The most common dissolve box, how- 
ever, is rigged up to handle only two channels. It is very simple to 
install and at least 80% of the dissolving or superimposing needs of 
field programs can be met if only two channels are available (Fig. 10). 

RCA now offers an "Auxiliary Field Switching Control" which 
operates in the same manner as their studio switching system de- 
scribed below. Since this is a separate box, it can be remote from 
the switching system if desired. Six channels may feed through this 
switching control which then feeds into the "Aux 5" button on the 
switching system. Five inputs are still usable on the switching 
system, in addition to the six that the auxiliary unit provides. A 
total of eleven local signals may be handled (six of which can be 
j^aded) or any combination of local and incoming remote signals may 
be used up to a maximum of seven remote and four local. This 
unit greatly expands the usefulness of the field switching system. 




The RCA Studio Switching System 

This equipment is often seen in both large and small studios. It is 
a dual-fading-bus type of switching system and is built into a unit 
with the master monitor. It is possible to remove the switching 
system from the monitor, however, and this has been done in some 
studios to allow either the T.D. or the producer to push buttons from 
the production desk. 

Fig. 11. Auxiliary field switching control. 

Fig. 12. RCA Studio Switching System. 

Basically, this switching system is built around two rows of but- 
tons, each feeding a fading bus. The two fading buses are controlled 
by levers which can either work separately to fade out one bus and 
fade in the other, or, as is more often the case, the levers can be 
clipped together to operate simultaneously for lap-dissolves. 

In Fig. 12, the two lower rows of buttons are connected to the two 

420 RUDY BRETZ April 

fading buses. The white row may be faded by the white handle, the 
black row by the black handle. Above the two rows of switches is a 
row of tally lights that show which channel is on the air. The tally 
light does not indicate which fading bus is operating; the position of 
the fading handles shows this. 

Straight switching between cameras can be done with either bank 
of push buttons, providing the fading control for that particular 
bank is open. The signal from Camera One, in other words, is car- 
ried to both Button One on the white row of buttons and to Button 
One on the black row. When White Button One is punched, the 
signal goes onto the white bus and through the white fader control, 
if that control is open. If a dissolve is desired to, say, Camera Two, 
Button Two must be preset on the black row (since the black fading 
control is closed, nothing happens to the program line). When the 
moment for the dissolve comes, the white bus is faded out (with 
White Button One depressed) and the black bus is faded in (with 
Black Button Two in the ON position). The two fading dials work 
in opposite directions; that is, the ON position for the white handle 
is at the top, and the ON position for the black handle is at the bot- 
tom. Thus at the beginning of the dissolve, both levers will be at 
the top, and so it is possible to work them with one hand and simul- 
taneously fade one picture out and fade the other in (there is a spring 
clip with which the levers can be fastened together). 

Superimposures. The halfway condition in a dissolve is a super- 
imposure, and the handles may be left in the midway position if de- 
sired. Each camera is then at half brilliance. This does not always 
make the best superimposure, however. Sometimes it may be neces- 
sary to superimpose a ghost onto a scene without making any notice- 
able change in the scene itself. The second camera, with the image 
of the ghost, must be added to the first, without lowering the level 
of the first signal. In this case the two handles are separated and 
the second one is brought up to the desired brilliance without lowering 
the first. Some operators like to keep their fading handles separated 
at all times, feeling that they thus have better control. 

Cutting to or from a Superimposure. The manufacturer of the RCA 
switching system says nothing about achieving this effect. It is not 
supposed to be possible, in a switching system composed of two fad- 
ing buses, to tie up both in a superimposure and then switch to an- 
other shot. There is nothing left to switch to; everything is in use. 
If another camera were punched up on either of the fading buses, the 


new picture would simply take the place of one of the other camera 
signals within the superimposure. One of the cameras must be lost 
from the superimposure before straight switching can be done. The 
same is true in getting into the superimposure. A straight shot must 
be taken first and the superimposed picture added; the two cannot 
appear at once. Technical directors with skill and ingenuity have 
discovered ways, however, to operate this equipment and effect a 
cut to a superimposure or away from a superimposure to a single shot. 

A practical problem like this sometimes arises: A girl enters a 
room, sees a ghost, and speaks to it. If the ghost were to materialize 
slowly before the girl's eyes, the problem would be easy. But the 
ghost must be in the room as soon as we see it. We must then cut 
from the room-and-ghost shot to a shot of the girl as she speaks. 

One method of doing this is as follows: Camera One takes the girl 
as she enters the door, Camera Two, the room, and Camera Three, 
the ghost (an actor in another set, brightly lighted against a black 
background). Camera One is on the air. If One is punched up on, 
say, the top row of buttons, punch it up also on the bottom row and 
set the handles halfway. Half of the Camera One signal is then 
coming through one fading bus, and half through the other. Now 
place one finger on Button Two in the top row and a second finger on 
Button Three in the bottom row, and press them both at once. Of 
course, there has been no way of previewing the superimposure. You 
are cutting blind, so to speak, and, unless you rely heavily on the 
cameramen to mark their finders during rehearsal and repeat the 
exact framing they had then, you will not be quite sure whether the 
ghost will seem to be standing next to or inside the furniture, with his 
feet on the floor, or floating several inches above it. To cut away 
from this superimposure back to Camera One again as the girl speaks, 
simply press both Camera One buttons at once. 

A second method of doing this is a little more difficult. With 
Camera One on the air as before, punched up on the top (white) row 
of buttons, a cut to a superimposure of Two and Three is desired. In 
this case, preset Black Button Two and hold a finger ready on Button 
Three in the white row. Both fading handles will be in the top posi- 
tion as shown in the illustration. At the same instant, press Button 
Three in the top row and bring the fading handles down halfway. To 
cut back again to Camera One by this method, simultaneously punch 
Button One in the top row and bring the fading handles to the top. 




RCA Program Director's Console 

This switching system is similar to the one just described except 
that it is built into a console containing three monitors, intercom 
controls, talk-back microphones and working table space for three. 
The table is for the T.D., the program director, and his assistant (Fig. 

An unusual feature of the monitors is that they are mounted verti- 
cally in the cabinet below the table and are viewed through a mirror. 

Fig. 13. RCA Program Director's Console. 


This permits the top of the console to be very low, making it possible 
for the director and the T.D. to see the camera control monitors and 
into the studio from a seated position. 

This equipment is used in a number of ways. Some studios mount 
the maximum of five monitors in the console, so that the director 
may have the three camera monitors in front of him, with a master 
monitor and a preview screen as well. 

The manufacturer intended the three screens to be used for: (1) 
preview; (2) studio switching master monitor; and (3) on-the-air 
master monitor. The difference between a studio switching master 
monitor and an on-the-air master monitor is a little obscure to the 
average production man. The engineer will use the studio switching 
monitor to show the results of operating the studio switching system, 
and the on-the-air monitor as a check on the condition of the picture 
farther on in the circuit and as a cue monitor. 

In actual practice, the two side screens are most often used as pre- 
view screens, but the exact use of the console is a matter of personal 
preference. Some directors at CBS use it only in connection with 
the regular camera control monitors, which they can watch down be- 
yond as in Fig. 1 (c). They use the left preview monitor only for 
film, the right only for still pictures. Even with three studio cam- 
eras, I have found it quite satisfactory to do the entire show without 
referring at all to the camera control monitors. The middle screen 
is a studio switching monitor. Camera One is always previewed on 
the left; Camera Three always on the right. When One is on the 
air, Two is found on the left monitor; when Three is on the air, 
Two is found over on the right. This is really not as confusing as it 
may at first seem, since each monitor has a series of numbered tally 
lights above it, and it requires but a quick glance to read what picture 
is displayed on each monitor. This previewing of camera channels 
is established as an automatic procedure to be followed by the T.D. 
without orders from the director. Channels other than these three 
camera channels can be previewed on either monitor at the director's 

Switching System. Switching allows for a maximum of 12 different 
inputs in this console. Small stations will not utilize all these posi- 
tions, but in a large network studio all may be needed. CBS, for 
instance, has five film channels in the projection room, and also a 
monoscope channel with test pattern. Any of these might be needed 
for a studio show, so that all must feed into each studio switching 




system. A possible maximum of four studio cameras (five have been 
used on complicated shows) brings the total number of channels to 
ten. One channel must be used for the incoming cue line, which is 
necessary in a network studio and shows the on-the-air network pro- 
gram at all times. It is previewed by the studio control room just 
before going on the air so that in case of an error in timing, the show 
can wait until the line is clear. All of this leaves only one channel 
for a spare (Fig. 14). 

The dual-fading-bus method of dissolving is used. One bank of 
buttons is white, and is controlled by a white fading handle; the 
other bank is black, with a black handle. The two fading controls 
work opposite to each other so that when they are moved sunultane- 



Fig. 14. Push-button panel used in RCA Program Director's Console. 

ously, one bus is faded out, the other in, and a dissolve is effected. 
This is almost identical with the RCA studio switching system. Two 
banks of push buttons at the top of the panel control the two preview 

CBS Adaptations from the RCA TC5A Switching System. Figure 15 
shows how CBS has simplified this switching panel, and added an- 
other row of buttons. Instead of having a row of tally lights above 
every row of buttons, they have used a plastic button which is itself 
a light, and lights up as soon as it is punched; hence the simplifica- 
tion. The added feature is a row of straight switching buttons not 
connected with the fading buses, and not controlled by either fading 
handle. Figure 6 shows the relationship of these buttons. 


The advantage of this is that one can cut to or from a superimposed 
effect. The last button (the 13th) on the row of straight switching 
buttons is labeled E, for "Effects." When that button is punched, 
the output of the fading buses goes out the program line. Thus it is 
possible to preset this entire effects system and preview the com- 
bined picture, and for this purpose an effects button is included in 
each row of preview buttons. To cut away from a superimposure, 
the next take is punched up on the straight switching bus, and the 
effect buses are dropped. 

The only disadvantage in this type of switching system is that the 
director must give the T.D. sufficient warning before calling for a 
dissolve. If the cuts are being made on the straight switching bus, 







Fig. 15. CBS adaptation from RCA push-button panel shown in Fig. 14. 

for instance, the T.D. has to bring the fading buses into operation 
before he can make a dissolve. To do this, he presets the same chan- 
nel that is on the air on one of the fading buses and punches the E 
button in his top line. No effect will be seen and he is then ready to 
preset the other fading bus for the channel he is to dissolve to, and to 
make the dissolve. This does make dissolving a little more difficult, 
but it may well be a good thing, in the light of the great amount of 
meaningless, unnecessary and disturbing dissolves used in television 
today. Dissolves have been much too easy to do. A switching sys- 
tem which sacrifices some of the freedom in dissolving for added flexi- 
bility in other directions is, it seems, built to a good design. 
An extra button has been added by CBS at the far left of each row 




This is the black button (no signal). It is easier to make a fade-out 
by dissolving to black rather than by fading out one channel. The 
two fading handles are thus kept working together, which eliminates 
certain possibilities for error. 

The Dumont Mixer 
Dumont manufactures only field equipment, but it is frequently 

Fig. 16. Dumont Mixer. 


used in the studio. The field switching system particularly is well 
adapted to studio purposes. It is a mixer type of switching system, 
with separate fading dials for each channel. The manufacturer 
labels it l 'Mixer- Amplifier and Monitor," in recognition of the other 
functions incorporated into the same suitcase unit (Fig. 16). 

Starting at the bottom of the mixer control panel, there is, first, a 
series of four tally lights for the four channels this equipment can 
handle. In the center of this row is an input socket for plugging in 
the T.D.'s earphones and talk-back circuit. Just above are the four 
fading dials, and above each of these is a push button switch. These 
four switches are mechanically interlocked so that if all faders are left 
open, straight switching can be done. 

An unusual feature of this switching system is an automatic fade 
and dissolve, which can be actuated by these same buttons. Above 
the fourth push button is a dial marked "Auto Fade Rate." If this 
dial is set at "Instantaneous," pushing the channel buttons results 
in straight switching. 

If the automatic fade dial is set on any of the other three positions 
(slow, medium or fast), a lap-dissolve or a fade-out-fade-in will be 
made. A small toggle switch to the left of this dial selects either the 
lap or the fade effect. The only manual control necessary is punch- 
ing the button for the new channel ; the dissolve or fade-out-fade-in 
then proceeds automatically at the selected speed. A fast lap is 
completed in Ij^ sec; medium speed is 3 sec; and a slow dissolve takes 
5 sec. The fade-out-fade-in proceeds at about the same speed as 
the dissolve. 

Manual Operation. Since the automatic dissolve cannot be halted 
at a halfway point for a superimposure effect, another method of oper- 
ation has been provided. In the middle of the row of .camera switch- 
ing buttons, divided from them by a white line, is a button marked 
"Manual Mixing." Punching this button throws all the other but- 
tons into the ON position. Control is then exercised only by means 
of the fading dials. 

During automatic switching all the faders must be open, but the 
switching buttons are interlocked, and only one channel at a time is 
switched into the mixer. When the manual button is punched, how- 
ever, all the channels are switched into the mixer and will all go out 
at once as one big superimposure, unless their fading dials are first 
turned to the OFF position. 

428 RUDY BRETZ April 

The transition from automatic to manual operation is usually made 
in this fashion: All the faders are turned down except the one con- 
trolling the channel which is on the air. The manual mixing button 
is then punched. There is no effect on the program picture, since all 
that the manual button has done is to turn on the other three switches. 
It is now possible to fade, lap or superimpose to any extent and at 
any speed, since everything is under manual control. If at any time 
a cut is desired, automatic cutting must be restored by punching the 
switching button of the channel that is on the air. No effect is seen, 
but all the other switches are then turned off. However, the faders 
also are off and must first be turned up before a switch between chan- 
nels can be achieved. 

Here is the routine again, applied to a simple example: Starting 
with Camera One, on the close-up of an actor, it is desired to cut to 
Two, on a long shot of the room, then superimpose a ghost with 
Camera Three, lose the superimposure and cut again to One. 

Director's Cue Technical Director's Action 

1. "Take Two." Checks to see that the automatic 

fader is set at INST. Punches 
Button Two. 

2. "Ready to superimpose." Turns down all dials except No. 

Two. Punches manual button. 

3. "Superimpose Three." Turns up No. Three fader. 

4. "Lose Three." Fades down No. Three fader. 

5. "Ready to take One." Punches up No. Two button. Turns 

% up all faders. 

6. "Take One." Punches No. One button. 

The method of operating this equipment usually is as follows: 
The right hand is kept on the automatic fader, ready to turn it to 
INST if the director calls for a take, or back to one of the other set- 
tings, according to the director's "ready" cue. The left hand is used 
for punching buttons and working dials on the first two channels. 
The right hand generally handles the third and fourth channels. 
If the director is not careful to give "ready" cues, however, he may 
call "Take" and get a dissolve (a very frequent phenomenon where 
this equipment is used) . In general, it may be said that the automatic 


dissolve feature increases the ease with which dissolves and fades can 
be made and leads to an over-use of these effects. 

Cutting to or from a Superimposure. While the direct cut into or out 
of a superimposure was not intended as part of the function of the 
Dumont mixer, it is possible to achieve this effect by one of two 

The simpler method is to punch two buttons simultaneously. The 
T.D. must mark the position of each fading dial during rehearsal, 
and set them thus before the switch so that the balance between 
them will be correct. It is often found desirable when two channels 
are punched up at the same time, to leave the fading dials full open 
and regulate the balance between the two pictures by riding the 
pedestal controls on the camera control units. This requires either 
very close co-ordination between the T.D. and the video engineer, or 
close proximity of mixer and camera control units so that the T.D. 
can operate both. Cutting from a superimposure to another camera 
is very simple when using this method; punching up the new camera 
will automatically release the previous two. 

A second method makes use of the manual button (which throws 
all switches into ON position at once). The T.D. opens the two fad- 
ing dials that control the cameras which are to be superimposed and 
then punches the manual button. The cameras will go on the line 
simultaneously. Automatic switching is no longer in effect,, how- 
ever, and the first switch will not automatically snap off. Unless the 
first camera is quickly faded down when the switch is made, there 
will be three picture signals on the air. 

To cut from the two superimposed cameras back to the single shot 
again, the T.D. simultaneously punches the button that controls 
this camera and turns up its dial. Punching the camera button 
throws the system into automatic switching and both the superim- 
posed cameras are dropped from the line as the new camera is switched 
in. Since the fading dial for this camera had been closed before, it 
must now be quickly opened so that the signal can feed through. 

To Delay a Fade-In. Sometimes a program will require a longer 
fade-out-fade-in than the automatic slow rate of 5 sec. In this case, 
the T.D. may punch the fade-out and quickly turn down the dial of 
the new channel so that the automatic fade-in will have no result. 
Then, when ready, he will fade in that channel manually by turning 
up the dial. Another way to do this is to make an automatic fade to 
black. Since the automatic fade effect always includes a fade-in 

430 RUDY BRETZ April 

after the fade-out, if the picture is to stay black, the new channel 
must be a dead channel. The T.D. will turn down the fading dial 
on a channel which is not in use, set the auto-fade dial and switch, and 
punch the button for that dead channel. The live channel will fade 
out, the dead channel will "fade in" and the screen will remain black. 
Then, when he is ready to fade in the next picture, he will punch the 
appropriate button. The mechanism then makes a second automatic 
fade; the dead channel automatically "fades out" and the new picture 
fades in. 

Fig. 17. General Electric Program Console. 

Dissolving to a Superimposure. This is an effect which cannot very 
easily be accomplished on any standard equipment except the Dumont 
mixer. At the mid-point of this effect three cameras must be on the 
air at once. When the system is set for manual mixing, it is possible 
to fade down a dial with one hand, and fade up two more simul- 
taneously with the other. This can be done, however, only if the 
two dials which are to be opened simultaneously are next to each 
other on the board. Technical directors with three hands are difficult 
to find. 


General Electric Program Console 

The switching system manufactured by General Electric is built 
into a console which seats three, the director, the T.D. and the audio 
engineer. Figure 17 shows this console in operation at WGN-TV in 
Chicago. General Electric has just placed on the market a new 
model which is part of its new line of television equipment. As far 
as its operation is concerned, however, it is reported not to differ 
greatly in any basic way from the console and switching system de- 
scribed herein. 

From the standpoint of production, this is a beautifully conceived 
piece of equipment. Everything has been thought of and built in, 
even to telephone handsets at each position and recessed ash trays. 
All that need be added is a script and the inevitable container of 

Monitors are not included in this console. It is intended for use 
with the General Electric Camera Control Desk, which includes 
large monitors, relatively high, so that it is not difficult for the director 
to watch them over the video operators' heads. The console can be 
used in the dark (a desirable condition where monitors are not hooded 
to keep off stray light) . Every control label is printed in fluorescent 
paint, which glows brightly when the console is flooded with ultraviolet 
(black) light. For the illumination of scripts, a small white light is 
built into the bottom of each gooseneck microphone mount, which is 
adjustable to any angle. 

The director's position on the left is equipped with a clock and built- 
in stop watch, plus a variety of intercommunication switches which 
can be hooked up to whatever combinations of circuits are desired. 
A disadvantage of this console is the lack of desk space for an assist- 
ant director. The assistant must work at an adjoining table or pull 
up a chair to the side and work without desk space. 

The audio operator's console at the right is equipped with a five- 
position mixer (five fading dials), with two outputs. Operating 
experience has shown that five positions are too few in television, 
especially in a large studio. Each source of sound requires a separate 
potentiometer in general practice; and although there are ways of 
circumventing this, they are not very satisfactory. A further dis- 
cussion of this point is to be found in the chapter on audio equipment. 

The switching system in the center of the program console is of 
primary interest at this point (Fig. 18). It is a dual-fading-bus type 

432 RUDY BKETZ April 

with the push buttons arranged in two banks, one for the left hand to 
operate, one for the right. Six inputs are provided. All of these may 
be used by cameras, or some may bring in film or remote signals. 
The method of operation is similar to that of the RCA studio switch- 
ing system. Straight switching is accomplished by punching but- 
tons in whichever fading bus happens to be in use. Two fading 
handles control the two fading buses. These handles are mechani- 
cally interlocked so that it is possible to make a dissolve by operating 
only one handle, the handle controlling the bus which is to be faded 
in. When this handle is faded in, the other automatically fades out. 
r [ he interconnection does not function, however, when the handle of 
the fading bus that is on the air is operated. As this bus is faded out, 
the other will not automatically fade in, which makes a fade-out to 




o a **AOV 

G^I O rfesi 9 



19 BJ 




|g>l w 

/fr- com 


Fig. 18. General Electric Program Console switching system. 

black possible. A superimposure can be made by stopping the handles 
halfway. Two channels cannot be superimposed at full strength, 
however, since bringing up the second handle automatically brings 
down the first one. This limitation makes the mechanical interlock 
feature a disadvantage in making superimposures. 

Above each push button is a series of three tally lights, the top one 
green and the other two white. The bottom white one lights when 
the channel is on the air, and shows which channel is on and which 
fading bus is in use. The middle row of white lights is marked 
" Available" and indicates the channel that has been preset. These 
lights work on the fading bus which is not in use but is preset for dis- 
solves. The green lights on the top row are preview or "ready" indi- 
cators. A light in this row shows that the T.D. has punched the 
proper one of a series of ready buttons which are placed on the top of 


the console on the extreme right. When one of these ready buttons 
is pushed, a green ready light goes on in the camera and on the camera 
control unit and, at the same time, the green tally light, which was 
just described, goes on in the switching system. The two top lights, 
then, should be on before a dissolve is made, after which the bottom 
light is then added. Verbal ready cues are very helpful, but this 
ready-light system seems to be an unnecessary complication and I 
have not seen it utilized very often in actual practice. 

Cutting to or from a Superimposure. There are two ways to do this 
on the General Electric switching system. The first method makes 
use of the built-in by-pass control. A special set of buttons (top row, 
far left) makes it possible to select any particular incoming signal 
and circuit it around the switching system. This is used, for instance, 
when a film is on the air (which would normally feed through the 
switching system), and a studio show must be rehearsed at the same 
time. It eliminates the extra switching panel which is used for this 
purpose with RCA equipment. This bank of by-pass buttons may'be 
used for straight switching if desired. 

Buttons on this bank control incoming program lines such as film 
channels and network line, and one button controls the output of the 
switching system. If the individual studio cameras are also fed 
through by-pass buttons, a camera may feed its signal to the trans- 
mitter without passing through either of the fading buses. The 
operator is then free to punch up two cameras on the two fading 
buses, set a superimposure, and cut to it by punching the switching 
system button on the by-pass bank. This button serves the same 
purpose as the E (Effects) button on the CBS modification of RCA's 
director's Console. The operator of the General Electric switching 
system may also preview this superimposure if he has a studio switch- 
ing monitor. Such a monitor shows only the output of the switching 
system, and is not a master monitor since it will not necessarily show 
the picture on the program line. A studio switching monitor in 
addition to a master monitor would be necessary in such a set-up: 
this is more than most stations which have installed this equipment 
have been able to provide. 

A second and more common method of cutting to a superimposure 
is as follows: One of the pictures to be superimposed is preset on the 
fading bus which is not in use. The other is switched onto the fading 
bus which is in operation and at the same instant the two fading 
handles are set halfway. 


There is a foolproof interlock on the two fading buses which makes 
it impossible to preset a camera on one bus if it is already in use on the 
other. This interlock may help avoid a few errors, but it renders 
impossible a method of cutting to a superimposure described in con- 
nection with the RCA studio switching system. One cannot cut to a 
superimposure by punching similar buttons on both sides, setting the 
fading handles halfway and punching two buttons at once. 

Each of the manufacturers designed the present switching equip- 
ment and production consoles two or more years ago. Each based 
its design on the observed production procedure at its own broad- 
casting studios; General Electric at WRGB in Schenectady, Dumont 
at WABD, and RCA at NBC. Production procedures vary from 
station to station and these differences between the manufacturers' 
testing grounds produced in large part the variation in the operating 
equipment just described. As new kinds of equipment are devised 
and tried out, a closer standardization can be expected. With the 
exception of patented features the future models of all three manu- 
facturers can be expected to approach a standard design. 

Color Temperature: 

Its Use in Color Photography 



Summary Color temperature as a specification for light sources is in- 
adequate to define any light source for color photography which departs ap- 
preciably in energy distribution from the black body. It should probably be 
restricted to use with tungsten incandescent lamps only. Meters devised 
to measure color temperature by means of measurements of the relative 
energy in two wavelength bands are likewise not trustworthy when applied 
to any but the black-body sources. A "three-point," rather than a "two- 
point," meter is needed for the precise control of photographic exposures. 
Such a meter should have sensitivity distributions that match those of the 
three emulsion layers of the color film. 

MANY important sources of light, such as the sun and the in- 
candescent lamp, belong to a class, sometimes referred to as tem- 
perature radiators, which emit light due to high temperature. The 
ideal temperature radiator, from which the concept of color tempera- 
ture is derived, is known as the black body, or complete radiator. It 
follows from the theory of the radiation from hot bodies, that a body 
which is a perfect absorber is also a perfect radiator. From theoreti- 
cal considerations it is possible to calculate the amount of light of any 
wavelength that will be emitted by a black body at any given tempera- 
ture. As the temperature is raised, an increased proportion of the 
energy is radiated at the shorter wavelengths, and the color changes 
from red through orange, yellow and white, to blue at a very high 
temperature. This series of colors forms the basis of the color tem- 
perature scale. It is important to note that many colors are not 
found on this scale and hence light sources having colors not matching 
the color of a black body cannot be expressed in color temperature. 
Examples are green, purple, magenta, violet, etc. 

Both the color and the energy distribution of a black body are 
known, once the temperature is specified; hence, when applied to 
practical sources, color temperature can refer to either or to both of 
these two aspects of the source. The definition of color temperature 
adopted by the Optical Society of America 1 refers to the color alone 

PRESENTED: October 11, 1949, at the SMPE Convention in Hollywood, Calif. 





> be 

I S 


and this definition has been generally accepted. 2 The color temper- 
ature of a light source may be defined as the temperature of a black 
body that matches the color of the source in question. 

It is a property of color vision that sources of many different energy 
distributions may appear exactly the same color. Furthermore, it 
can be demonstrated easily that the suitability of a source for the 
illumination of colored objects is determined more by its energy dis- 
tribution than by its color. Hence, unless the source has a known 
energy distribution, the color temperature specification is not usually 
sufficient to describe the most important aspect of the source. The 
energy distribution curves of Fig. 1 are characteristic of those of a 
black body. The temperatures are expressed in the absolute scale, 
in degrees Kelvin (C -f 273). Many familiar practical sources of 
light such as the tungsten incandescent lamp, not only match the 
color of a black body, but even have nearly the same energy distribu- 
tion, although the actual temperature may differ from the color 
temperature. A tungsten filament, for example, may have a color 
temperature of 3000 K when its actual temperature is 2970 K. The 
candle flame is another example of a light source which closely re- 
sembles a black body both in color and in energy distribution. Such 
sources may be specified sufficiently well for most purposes in terms 
of color temperature alone because the energy distribution is known 
to be similar to that of the corresponding black-body source. 

Many other sources, however, have energy distributions that are 
very different from those of a black body, although they may match 
the color of a black body at some temperature. An example is the 
fluorescent lamp. Illustrated in Fig. 2 is the energy distribution of a 
fluorescent lamp known to the trade as "3500 White," compared to 
that of a tungsten lamp operating at a color temperature of 3500 K. 
While these two sources in themselves have nearly the same color, 
the colors of familiar objects often look quite different under them. 
These differences in appearance are due to the differences in the en- 
ergy distributions. The practical problems created by such differ- 
ences in energy distribution are recognized by meat dealers, for ex- 
ample, who have observed the unfavorable dark red appearance of 
choice meat when illuminated by fluorescent lamps. Diners in res- 
taurants also have objected to the greenish appearance of egg 

If the source departs radically from the energy distribution of a 
black body, then a color temperature specification is of questionable 






u. o 


_J * 



1= I 


< m 


value. The color temperatures of the two sources illustrated in Fig. 
2 are the same; so color temperature in this case fails to distinguish 
between a good source and one which is much less suitable for the il- 
lumination of certain colored objects. Some authorities 3 ' 4 - 5 are al- 
ready giving serious consideration to the complete abandonment of 
color temperature as a specification of light sources, or at least re- 
striction of the use of such specifications to tungsten incandescent 
lamps alone. Jones 3 suggests the use of the true temperature as an 
index of the spectral distribution of radiant energy from tungsten 
lamps. The important point to be recognized is that color tempera- 
ture is a color specification, and as such it is inadequate to describe 
any but a very restricted class of artificial sources, principally tung- 
sten lamps. It is inadequate because it fails to describe the most 
important aspect of the source, its energy distribution. 

The inadequacy of a color specification for light sources may be 
further illustrated by an even more striking example of the effect of 
the energy distribution of an illuminant on the appearance of a colored 
object. While we usually consider white light to consist of a mixture 
of light of all the wavelengths of the visible spectrum, it is a well- 
known property of color vision that white also can be simulated by a 
mixture in the proper proportion of lights of only two, or at most, 
three wavelengths. Two such wavelengths are called complementary 
wavelengths. Consider two sources, then, with energy distributions 
such as those illustrated in Fig. 3. Both of these sources would look 
substantially white to a normal observer. They have the same color 
temperature. Source A is the standard I.C.I. (International Com- 
mission on Illumination) Illuminant C, and is similar to one phase of 
daylight; hence, objects appear in their normal colors under source 
A. Source B, however, contains light of only two wavelengths, one 
in the yellow, the other in the blue. Let us consider, now, the dif- 
ferences in the appearance of some typical colored objects under 
these two sources. 

The color of an object results from the selective reflection, absorp- 
tion or transmission of the light of various wavelengths falling on its 
surface. Thus a red object may reflect only the red wavelengths and 
absorb light of all the remaining wavelengths. This property of a 
surface may be represented by a curve such as that in Fig. 4, which 
shows the percent of the incident light that is reflected by the surface 
of a red object at each wavelength of the spectrum. When the light 
from any source falls on a colored surface, it gives rise to reflected 





* ro 




light which generally has an energy distribution different from that of 
the source. It is the energy distribution reaching the eye from the 
object that defines the resulting color stimulus. This energy distri- 
bution may be obtained for the red object under source A by multi- 
plying the ordinates of the curve in Fig. 4 by the ordinates of the 
energy distribution curve for source A in Fig. 3. The resulting en- 
ergy distribution reaching the eye from the red object is shown in Fig. 
5. Since only the red wavelengths are reflected, the object appears 
red. For source B, however, there is no red light from the source but 
only blue and yellow, both of which are absorbed by the red object, 
so no light at all is reflected and the red object appears black. By 
similar reasoning, it is possible to show that with the same source 
other objects may appear yellow, blue or neutral, depending on their 
relative reflectances for blue and yellow light. Figure 6 is the re- 
flectance curve of a green object. Since the reflectances of this object 
for the two lines of source B are equal, the object appears the color 
of the source or neutral gray. While our example, source B, repre- 
sents ah improbable extreme, it differs in degree only from many 
practical sources. Some fluorescent lamps have a deficiency of en- 
ergy in the red region which is balanced by a corresponding deficiency 
in the complementary wavelengths, the blue-green. Also there is an 
excess of energy in the yellow and green which is balanced by an ex- 
cess in the complementary wavelengths, the violet. Lamp manufac- 
turers are aware of these deficiencies and are working steadily for the 
improvement of commercial fluorescent lamps. 

Commercial processes of three-color photography employ films 
having three sensitive emulsion layers with spectral sensitivities some- 
what similar to the sensitivities of the color receptors of the eye. 
Hence color photography is almost exactly analogous to color vision 
in this respect. Color adaptation, by which familiar objects tend to 
retain the same appearance whether viewed by tungsten light or day- 
light, is, in the end result, analogous to the use of color correction 
filters with the color film so that objects photograph the same color 
by different light sources. And with color photography, the energy 
distribution of the source is just as important a factor in the appear- 
ance of objects as it is in vision. 

It is customary, because of the convenience, to specify tungsten 
studio lamps in terms of the color temperature for which the film is 
balanced. Thus, Kodachrome Professional Film Type B requires a 
color temperature of 3200 K. Lamp manufacturers supply lamps 

442 0. E. MILLER April 

that will operate at approximately this color temperature when burned 
at their rated voltage. Since tungsten lamps change in color with 
ago, changes in line voltage and blackening of the glass envelope, color 
photographers have had a real need for some means of checking the 
color temperature of their lamps. To supply this need, color tem- 
perature meters are now available. If voltage control is provided 
for in the studio, a color temperature meter offers a guide for the ad- 
justment of the lamp voltage to obtain the proper color temperature. 
If the lamp voltage cannot be adjusted, the meter reading may indi- 
cate when a color correction filter is needed. 

Several color temperature meters are designed to measure color 
temperature in terms of the ratio of the energies in two separate 
wavelength regions of the spectrum and are calibrated either to read 
color temperature directly, or to read in terms of the color filter re- 
quired. It is assumed that the energy distribution is of the black- 
body type, and hence the readings are liable to serious error if the dis- 
tribution departs seriously.from that assumed. This limits the useful- 
ness of such meters almost entirely to tungsten incandescent lamps. 
Unfortunately, some of these meters may be used with daylight. 
But a color temperature specification of daylight is difficult, because 
ttye energy distribution of daylight departs considerably from that of a 
black body. This does not mean that daylight is an unsuitable source 
for color photography, but rather that it does not match a black body 
in color and therefore cannot be specified accurately in terms of color 
temperature. Daylight is made up of a mixture of varying propor- 
tions of sunlight and skylight. While sunlight alone, or skylight alone, 
are rough approximations to black-body radiation within the visible 
spectrum, mixtures of the two are not, since an additive mixture of 
the radiation from two black bodies of different temperatures will not 
match a black body at any temperature. Figure 7 illustrates this 
point by showing one mixture of the radiation from two black bodies 
operating at 2000 K and 20,000 K, respectively. A mixture somewhat 
resembling this might be encountered in late afternoon when the sun 
is quite red and the sky a bright blue. Also shown is the energy dis- 
tribution of a black body at 3000 K. A color temperature meter 
based on measurements of the energies at wavelengths 520 and 690 
mju would indicate that the mixture was equivalent to the black body 
at 3000 K, whereas it has much more energy in the blue and red and 
less energy in the, green than the 3000 K black body. The mixture is 




quite pink and very far from matching a black body at any tempera- 

In view of the departures of daylight from black-body radiation, 
which are neither constant nor systematic, there is some room for 
doubt as to the usefulness of a color temperature specification in con- 
nection with daylight, and existing color temperature meters may 


100 - 


500 600 

Wavelength m>t 

Figure 7. 

give very misleading results. For many purposes satisfactory expo- 
sures can be made in daylight without any color compensation, pro- 
vided exposures are avoided during the hours of early morning and 
late afternoon. At such times results are often unsatisfactory any- 
way unless the object is to record the sunset or sunrise effect itself. 

In some applications of color photography precise control of day- 
light exposures may require that occasional filter corrections be made 

444 O. E. MILLER 

for the normal variations in the quality of daylight. Since it is the 
relative exposure of the red, green and blue records in the film that 
determines the color balance of a picture, a meter is needed which is 
capable of measuring the amounts of energy in the red, green and 
blue regions of the spectrum. Ideally, such a meter would have three 
sensitive elements with spectral sensitivities like those of the three 
emulsion layers of the color film. The three readings could then be 
interpreted in terms of the exact filter corrections needed .to give 
a balanced exposure. For such a meter to be useful a minimum of 
computation should be required to translate the meter readings into 
filter corrections. 


(1) S. M. Newhall and Josephine G. Brennan, "Comparative list of color 
terms," ISCC Report, p. 17; January, 1949. 

(2) American standard illuminating engineering nomenclature and photo- 
metric standards (Z7. 1-1 942), American Standards Assn., New York. 

(3) L. A. Jones and H. R. Condit, "Photographic exposure, II," /. Opt. Soc. 
Amer., vol. 39, pp. 112-123; February, 1949. 

(4) P. Moon, The Scientific Basis of Illuminating Engineering, pp. 130-132; 
McGraw-Hill, New York, 1936. 

(5) R. M. Evans, An Introduction to Color, pp. 212-214; John Wiley, New 
York, 1948. 

An Experimental 35-Mm Multilayer 
Stripping Negative Film 



Summary The so-called "three-strip" method is generally considered 
to give the best results when taking professional color motion pictures. 
An objection to this method is that it requires the use of a special camera 
fitted with a beam-splitter prism. This paper describes work that led to 
the development of a single multilayer negative film which can be used in 
any standard motion picture camera. After exposure and before develop- 
ment, the two upper layers are wet-stripped separately onto special transfer 
supports. Thus the single support of the multilayer film bears three color- 
sensitive emulsions in the following order: a red-sensitive layer next to 
the support, an intermediate green-sensitive layer, and an upper blue- 
sensitive layer. The red- and green-sensitive layers, and the green- and 
blue-sensitive layers are each separated by interlayers which facilitate 
stripping of the two upper layers. The film has an over-all thickness about 
the same as standard motion picture negative film. Its speed is such that 
good negatives of an open landscape can be exposed at / 8 or even //1 1. 
An experimental stripping machine is described which accurately registers 
the perforations of the two stripped layers with those of the original film on 
which remains the red-sensitive layer. . 

IT is GENERALLY ACKNOWLEDGED in the motion picture industry 
that the method of taking professional color motion pictures that 
has given the best all-around results is the so-called " three-strip" 
method. The separate films, after exposure, are developed in a 
regular negative developing machine, resulting in three excellent 
original tricolor negatives. An objection to this method is that a 
specially designed beam-splitting camera must be used. The question 
came to mind: Would it be possible and practical to design a single 
film which could be used in any standard motion picture camera, with 
regular optics, and which, after processing, would give a set of original 
tricolor negatives of quality comparable to those made by the three- 
strip method? One answer seemed to be a stripping film. 

Late in 1940 experiments were started on the design of such a 
film on which a single support would bear three color-sensitive emul- 
sions, the first layer next to the support being red-sensitive, the next 
layer green-sensitive, and the top emulsion blue-sensitive. Between 

PRESENTED: October 10, 1949, at the SMPE Convention in Hollywood. 
Communication No. 1312 from the Kodak Research Laboratories. 





the red-sensitive layer and the green-sensitive layer would be a special 
interlayer having the property of adhering to the emulsions when dry 
but readily coming apart when wetted in water. It was thought that 
hydrolyzed cellulose acetate of a suitable acetyl content would prob- 
ably meet these requirements. An interlayer of the same type would 
separate the green-sensitive emulsion and the blue-sensitive emulsion 
layers. Because of the fact that the green-sensitive layer and the 

Blue-Sensitive Emulsion 

Yellow Filter Layer 

Green-Sensitive Emulsion 


Red-Sensitive Emulsion 

> Sub. 


Fig. 1. Cross section of multilayer stripping film showing relative position 
of emulsion layers, filter layer and stripping interlayers. 

Fig. 2. Sample negatives 

obtained with early 


red-sensitive layer are also sensitive to blue light, it would be neces- 
sary to interpose a yellow filter layer between the blue emulsion and 
the green emulsion (Fig. 1). 

In the first experiments, a two-layer film only was attempted, since 
obviously if a single stripping operation could not be done success- 
fully, satisfactorily stripping the more complicated triple layer would 
be highly improbable. 


On February 6, 1941, exposures of a resolving-power chart were 
made in a Mitchell camera on such a two-layer film. The stripping 
was done before development on a simple machine on which the strip- 
pable emulsion layer was transferred onto a special transfer film 
bearing a suitable substratum. There was no pretense at maintain- 
ing registry between the two images, which, of course, must be accom- 
plished for motion picture making. Figure 2 shows actual samples 
of these first negatives. 

The technique of stripping was simple. The two-layer film was 
wetted in plain water, temperature approximately 70 F, for about 
10 sec, and while under water was brought in contact with the trans- 
fer film; then the two films were rolled into intimate contact between 
two rubber-covered rollers. The "sandwich" was then allowed to 
remain for about a minute in order that the emulsion layer about to be 
stripped could bond to the transfer film. The actual stripping 
operation was performed over two rollers. After drying the film, 
development was performed in the conventional way. 

About this time a blue-sensitive emulsion, bearing on its surface a 
yellow filter layer, was coated on a separate film. This made bipack 
experiments possible. Quite a bit of this bipack work was done on 
4 x 5 in. and 5 x 7 in. sheet film, the bipack being exposed in a camera 
having a sheet of plain glass in the film holder. Some useful experi- 
ments were carried out in this way insofar as the emulsion properties 
were concerned, but the scheme was soon abandoned as far as serious 
picture making might go, because of the many defects in the nega- 
tives, such as Newton rings, dirt and halation. 

Soon after this, a crude 35-mm machine was assembled which per- 
mitted stripping some film in which the perforations of the two-layer 
film were held in precise registry with the transfer film during the 
bonding period. 

Color prints from these two negatives were made on a contact 
registering printer. The prints were found to be in good registry 
and no signs of distortion of the stripped emulsion layer were evident. 

After this encouragement, a three-color multilayer film was at- 
tempted, but it was not until July 29, 1942, that a successful three- 
color coating was available for camera tests. The over-all thickness 
of this multilayer film was approximately the same as a black-and- 
white motion picture negative film. The sensitometric characteris- 
tics were excellent, all three emulsion layers having an exposure lati- 




Gommo Fog 
.52 .06 +Groy Base 
.52 .04 

.52 .02 (After Intensification) 
19 (Before Intensification) 


Log E 

Fig. 3. Typical sensitometric curves for the three layers 
of multilayer stripping film. 

Fig. 4. A typical set of multilayer stripping film negatives. 




tude comparable to that of regular negative material (Fig. 3). The 
speed was such that good negatives of an open landscape could be 
made at //8 or even //ll. The keeping properties of the emulsion 
layers equaled those of regular black-and-white emulsions. Because 
of the thinness of the three emulsion layers, the gammas were on the 
low side, the red- and green-sensitive layers giving a gamma of about 
0.55, and the blue-sensitive layer still lower, being approximately 

Fig. 5. Schematic section of machine Fig. 6. Schematic section of machine 
showing rolldown registering station. showing stripping station. 

0.30. The blue-sensitive layer was purposely made with a low silver 
halide density in order to interfere as little as possible with the resolv- 
ing power of the two underlying emulsion layers. For most printing 
processes this necessitated either intensifying the blue negative or 
making a "dupe" having the desired over-all gamma. 

When working out a motion picture film, it was considered impor- 
tant to succeed in stripping prior to development in order to avoid 
defects arising from depth development. As is well known to those 





working with material coated in several layers, these defects can be 
quite serious. The defects are caused mainly by the reaction prod- 
ucts in the lower layers diffusing to the upper layers. Another ad- 
vantage in stripping before processing is that a certain measure of 
gamma control is possible by regulating the amount of development 
in the usual way. 

Good progress was being made both in the manufacturing end and 
in the processing end. Many thousand feet of tricolor film were 
stripped, developed and printed, although the process could not yet 
be regarded as ready for the market. The photographic quality 
of the negatives was of a high order of excellence (Fig. 4) and ob- 
viously could be used in any of the known color printing processes. 

During the war years, work on the preparation of coatings was 
practically brought to a standstill, though worth-while efforts were 
continued in other directions, particularly in the design of a register- 
ing-stripping machine. After two or three machine designs had been 
tested and rejected, the basic principles on which a practical machine 
could be built were established and practical experimental work was 
done as circumstances permitted, ending in the construction of a 
machine, the essential parts of which are shown schematically 
in Fig. 5. 

In the figure, A is a tank of 70 F water in which the multilayer 
stripping film and the transfer film are wetted for approximately 10 
sec. The two films meet at the rubber-covered wringer rollers, B y 
where they are rolled into intimate contact. Four inches farther on, 
the perforations of the two films are brought into accurate registry 
by means of a specially designed sprocket, C, and a socket roller, D. 
The sprocket roller, C, is positively driven at the film speed of 30 
fpm. Since the perforation pitches on any two films are never 
exactly alike, it is necessary on the machine to keep the shorter-pitched 
film stretched to match the pitch of the longer one during at least part 
of the bonding time. This is accomplished by means of two other 
sprocket-socket roller assemblies, one of which is indicated in E. 
These two sprockets are friction clutch-driven and have an overdrive 
of approximately 5%, thus maintaining the required tension. The 
distance between the registering sprocket and the first overdriven 
sprocket is 18 in., and it is 48 in. between the first overdriven and the 
second overdriven sprocket. 

By the time the film arrives at the second overdriven member, the 
adhesion between the emulsion and the transfer film is such that the 




Fig. 8. Hand-stripping of multilayer film. 

Fig. 9. Close-up of stripping machine showing rolldown registering station. 


films can continue on their path without being under tension, al- 
though the bonding is not firm enough for stripping. 

In Fig. 6 the "sandwich" next is seen passing over a sheave, F, 
then down into a loop, and finally up to the two simple stripping 
rollers, G. After parting company, the two films pass through a 
drying cabinet of conventional type from which the stripped emulsion 
passes on to a regular take-up at the end of the machine (Fig. 7), 
while the red-plus-green multilayer film, when dried, continues on to a 
second wetting, rolldown, registering station, and is stripped in the 
same manner as the blue-emulsion layer. The red-sensitive emulsion 
remains on the original support. After these two films are dried in a 
second drying cabinet, they are also taken up at the end of the ma- 
chine. Development is done on another machine. 

In Fig. 8 a small piece of film is shown being stripped by hand. 
Figure 9 is a close-up photograph showing in detail the important 
rolldown registering station on the stripping machine that was 
sketched in Fig. 5. In the photograph (Fig. 9) is seen a straight film 
track. It is important that the films, when registered, move on 
through the machine in a straight line until the adhesion is sufficient 
to prevent slippage occurring between the two films. As previously 
mentioned, this state is reached immediately after the film has reached 
the second overdriven sprocket. The track serves to maintain the 
film in the required straight path and also acts as a means of stripping 
the films from the registering sprocket. The purpose of many of the 
other details in the photograph will be easily understood by film- 
processing engineers. 

NOTE: At the .conclusion of the paper, a short motion picture was shown which 
demonstrated the method of stripping the multilayer film, first by hand using 
small pieces, then by the registering and stripping machine with long lengths of 

Printing Equipment 
For Ansco Color Film 


Summary The Scenetester takes the place of a cinex machine in manu- 
facturing Ansco Color prints. Its primary function is to make it possible for 
the timer to select for a given color original or dupe the proper color correc- 
tion filters and printing density. The Scenetester must match the printer or 
printers it is teamed with for absolute light output, color balance, exposure 
time and light changes. This paper describes a Bell & Howell Model D 
printer modified for color printing and a Scenetester which was made for 
darkroom operation. The Scenetester prints 16 frames simultaneously 
through 16 different color correction filters and was designed to work with 
modified Bell & Howell Model D printers. 

THE PRINTER MODIFICATION consists of adding a new lamphouse, 
and the installation of an automatic filter changer (Fig. 1). The 
filter changer works on the principle of a slide projector. The indi- 
vidual filter packs are mounted in semitransparent filter holders and 
are stacked in the proper sequence in a feed magazine. The con- 
tactor originally operating the semiautomatic light-change mecha- 
nism and light-card indexer is employed in this modification to acti- 
vate also the solenoid of the filter changer. For every light change, 
a filter slide change automatically takes place, whether or not a 
filter change is necessary. 

Figure 2 shows the optical system employed. The light source is 
a standard 750-w, 120-v, T-12 pre-focus projection-type lamp. In 
operation this lamp voltage is adjusted to approximately 95 v, giving 
it a color temperature of about 2900 K (degrees Kelvin). This low 
color temperature increases "the life from the rated 50 hr to about 
1600 hr. 

A condenser focuses the filament of the lamp in an objective lens. 
The objective lens in turn focuses a plane in the condenser on the 
ground glass of the Bell & Howell printer, located at the aperture of 
the light-change mechanism. 

A fire shutter is located between the lamp and the condenser. A 

PRESENTED: October 11, 1949, at the SMPE Convention in Hollywood. 


filter holder is located on the other side of the condenser in the least 
intense part of the light beam. This filter holder accommodates the 
emulsion correction filters, an Aklo heat absorbing glass, and an 
Ansco UV-16 printing filter. Part of the forced air for cooling the 
printing light is directed over this filter pack. 


The basic elements of the Scenetester are a light source, optical sys- 
tem, a calibrated light-change mechanism, a curved platen for 16 
color correction filters, means for holding the camera film over the 
platen and a printing film stock carriage. The lamp receives its 
power directly from an a-c line through a voltage regulating trans- 
former. The color temperature of the light is set by a variable trans- 
former (variac), with an a-c voltmeter connected directly across the 
terminals of the lamp. Forced air ventilation is provided for the lamp 
and the emulsion color correction filters to permit continuous opera- 
tion of the lamp. 

The optical system of the instrument is similar to that of the 
printer modification, particularly the light source and lens system 
(see Fig. 3). A rotating mirror, mounted on a shaft driven by a 
synchronous motor and centered with the axis of the lens system, de- 
flects the light beam from the objective lens through an angle of 90, 
and causes it to scan the curved printing platen. The lengthwise 
centerlme of the platen aperture is aligned with the axis of the rotating 

The mirror is fastened inside a tunnel to keep stray light from fall- 
ing onto the film. An adjustable aperture is located on the end of 
the tunnel within J| in. of the camera stock being tested. This aper- 
ture may be adjusted to give any exposure between J4 and J^o 
sec to match the exposure time of the printer \vith which it is 
teamed. With a Bell & Howell printer running at 60 ft per min, the 
exposure time is set for J^g sec. 

The length of the platen opening is slightly over 12 in., accommodat- 
ing 16 single-frame color correction filter combinations. The radius 
of the platen is such that the camera and printing films are engaged 
over an arc of 140. 

Two exposure platens were made for each instrument. One of the 
platens was furnished with the color correction filter packs shown 
beside Fig. 4, each filter combination covering one 35-mm frame. 
Such a platen is now being used for making scene tests. 







Color correction filters come in three series : 

1. The yellow series (20), which absorbs blue and transmits green 
and red. 

2. The magenta series (30), which absorbs green and transmits 
blue and red. 

I I I 


Fig. 2. Modified Bell & Howell 
Model D printer optical scheme. 

Fig. 3. Scene tester scheme. 

Frame Filter 

1 None 

Frame Filter 

11 1-24 












































































Fig. 4. Ansco color 
correction filters. 





3. The cyan series (40), which absorbs the red and transmits blue 
and green. 

The numbering system is arranged so that 4 (i.e., 24, 34 and 44) 
will give a reduction of one-half stop of light in printing value for Ansco 
Color printing films in the desired band. Numbers 3 denote Y^ 5 
denotes 1, and 6 denotes 2 stops. 

The curves in Fig. 4 show the insertion losses of these filters plotted 
in density versus wavelength. 

From these curves it is evident that the least number of filters 
will give highest efficiency. For instance, one 24 will give J^ stop 
correction in the blue end with only 15% insertion loss in the rest of 
the spectrum. Two 23 's will give also J^ stop correction but will 
increase the insertion loss in the rest of the spectrum to about 27J^%. 
With the amount of light used in printing color film this becomes a 
factor to be reckoned with. Filters of all three series should never 
be put into one pack because, for instance, one 24, one 34 and one 44 
would act the same as putting a neutral density of approximately 
0.15 into the light beam; however, one 26, one 36 and one 45 can be 
replaced by one 25 and one 35 with one stop more light available to 
the film. 

The other platen was furnished with two methyl methacrylate 
resin (Lucite) window leaves J^g in. thick. Sufficient room was 
allowed in the construction of the platen to place a neutral dye or silver 
step wedge between the two pieces of plastic. This platen is used 
to make intensity scale sensitometric strips. 

The scene test and sensitometric platens are interchangeable with- 
out the use of tools during darkroom operation. 

A filter holder for the emulsion correction filters, Aklo glass, and re- 
lated items, is provided as in the printer. It is accessible during 
darkroom operation. Filters can be changed while the printing light 
is burning without fogging the raw stock. 

The light-change aperture is located between the elements of the 
objective lens. It has 21 discrete steps, being separately adjustable 
to give a light change from J^ 2 - to M~stop per step. The total range 
for the 21 steps can be adjusted to give from 1% to 4J stops variation 
from light 1 to 21. To assure that the flatness of field at the film is 
not altered when changing the light intensity, a ground glass was 
placed between the lamp and the condenser. The ground glass also 
makes placement of the filament of the lamp less critical. The illumi- 
nation across the film width is uniform within 2%, while the illumi- 

460 F. P. HERRNFELD April 

nation along the platen length is within = fc 3%. This uniformity was 
verified by film exposures. 

The camera film to be tested is held between two rewinds. It is 
threaded through a viewing box, over the film platen and to a take-up. 
The viewing box is located to the left of the film platen, with light 
shields to permit darkroom operation. The primary function of the 
viewing box is to frame the camera original with the color correction 
filters in the platen. Sprockets, rollers and weighted rollers are ar- 
ranged in such a manner as to allow easy threading in the darkroom 
with a minimum of danger to the camera film. 

The printing film is mounted on a manually operated carriage which 
accommodates 400 ft of Ansco Color printing film raw stock (see 
Fig. 5) . The spindles holding the film are made to take either nega- 
tive or positive plastic film cores. The film is fed from the unexposed 
roll over a film sprocket, under a curved pressure plate, over another 
film sprocket, and onto the take-up spool. The two film sprockets 
are coupled by a chain-and-sprocket arrangement to assure a con- 
stant loop under the pressure plate. When the carriage is moved 
down into the exposure position, i.e., at the point of contact between 
camera and printing film, a microswitch starts the exposure cycle. 
This cycle is completely automatic and has safety features preventing 
the exposure of less than a full platen, as well as double exposures. 
A contactor in series with the fire shutter also prevents accidental 
fogging of the film due to too early release of the carriage. 

As the camera and the printing film are engaged over an arc, it is 
necessary to move the printing film vertically away from the camera 
film until the loop of printing film has cleared the camera film by at 
least % in. before the raw stock is advanced. Only after all physical 
contact between camera and printing film has ceased, will the auto- 
matic transfer of the printing film start. At each upward stroke of 
the raw stock carriage, at least 20 frames of printing film are trans- 
ported onto the take-up spool. At the completion of the upward 
stroke, a contactor automatically resets the instrument for the next 
exposure cycle. If the upward stroke is not completed the fire shutter 
will not open upon contact of the printing and camera film. 


In adjusting the Scenetester for the proper light and color tempera- 
ture to match the color printer, a color temperature meter of the com- 
parative type is an invaluable aid (see Fig. 6) . A meter built for this 




particular purpose utilizes a Wcston Model 856 photovoltaic cell and 
a 200 microammeter having resistance of approximately 12 ohms. 
With this low resistance the meter reading will be practically linear 
with change in light intensity. Approximately 75 ft-c will give 
basically full scale deflection. 

Fig. 6. Color temperature meter. 

Fig. 7. Relative sensi- 
tivity of color meter 
and niters. 

S %L> 

TOM 8J6_ 

500 600 


A neutral density and two filters are used in conjunction with the 
meter. The first is a fine grain silver density to reduce the light in- 
put to the photovoltaic cell by a factor of 5. The second filter is a 
blue-green Corning glass #5031, 5 mm thick. The third is a red Corn- 
ing glass #2404, 2 mm thick. The transmission characteristics of the 
two color filters are shown in Fig. 7. 

462 F. P. HERRNFELD April 

The filters are mounted in a slide holder which fits the gate casting 
of a Bell & Howell printer and which is provided with an adapter for 
use with the Scenetester. The aperture of the meter is made to take 
the full light output from either the color printer or the Scenetester. 
This slide holder has four positions. In the first position the light 
reaches the photovoltaic cell directly, allowing the meter to give a 
readable deflection at the lowest printer point. In the second posi- 
tion the opening is covered with the silver density reducing the sensi- 
tivity of the meter from 75 to about 375 ft-c. In the third position 
the opening is covered with the blue-green filter. With the slide in 
this and the next positions, a T-pad is switched into the electric cir- 
cuit of the meter. This pad is for the purpose of adjusting the meter 
to 100%, the reference point for color temperature measurement. 
The fourth opening is covered with the red filter. The difference in 
reading between the two filters is the color temperature indication. 
The electric circuit of this color and light meter is shown in Fig. 8. 

FA Clear PVC Weston 856 cell 

FB 0.70 neutral density Pi 15-ohm T-pad 

Fc Corning 2404, 2-mm filter Si D.P.D.T. switch 

FD Corning 5031, 5-mm filter 
Mi Weston 200 U.A.D.C. meter 

Fig. 8. Color meter scheme. 

All measurements on the color printer and the Scenetester must be 
made with the heat absorbing glass and the Ansco UV-16 printing 
filter in place. 

In practice the lamp rheostat of the color printer is adjusted until 
the light meter reads between 240 and 350 ft-c with the printer set to 
picture aperture and maximum light. After this adjustment the 
pad is set to give a meter reading of 100% with the blue-green filter. 
Then the red reading is noted. The meter should read about 80% in 
reference to the blue-green. If the red reading is higher than 80%, 
the lamp voltage is increased and the light output decreased by insert- 
ing some neutral densities such as a fine mesh screen, a ground glass, 
or for smaller changes, clear glass, into the optical system of the 
printer. If the red reading is lower than 80%, the lamp voltage 
should be lowered. 

Measurements on the Scenetester are inverted. With a heat ab- 
sorbing glass and an Ansco UV-16 filter in place, the color tempera- 


ture is adjusted first and then the necessary changes are made to ob- 
tain the same light output as that of the color printer. 

All measurements made with the light meter should be considered 
preliminary. The light-change mechanism should not be calibrated 
with the meter. The final check at about every third printer light 
must be made photographically. For this a test loop is made. The 
picture of the test loop should consist of a neutral gray scale. Each 
gray patch should be large enough to be measured with a color densi- 
tometer. This loop should then be printed on the printer and Scene- 
tester every third light. From this test the light-change mechanism 
of the Scenetester should get its final adjustment. One or two re- 
prints may be necessary before a complete agreement between printer 
and Scenetester is reached. 


Two of the Scenetesters have been completed and have been in con- 
tinuous use since they were finished, one for a period of over two years. 
Both of these units are teamed with modified Bell & Howell Model D 
printers. Thus far they have performed well with very little main- 
tenance. They are checked on a daily routine basis for intensity and 
color temperature with the color meter. 

16-Mm Film Color Compensation 



Summary Second generation color duplicates on 16-mm film are noted 
for problems of contrast and color fidelity. These deficiencies have been 
responsible for the widespread practice of printing from the original. Some 
experimental techniques and devices for the printing of special key* inter- 
mediates which tend to counteract these faults are explained. 

DURING 1948, the greater part of 16-mm color release printing at 
the National Film Board of Canada was on Kodachrome du- 
plicating stock printed from originals. Some productions were printed 
from masters, but nearly all such second generation material far ex- 
ceeds the acceptable limits of contrast gain. Even duplicates printed 
from the original are seldom ideal in this regard. 

The nature of documentary film shooting makes it impossible to 
assume that a given original is a standard to which all duplicates 
should closely correspond. Because of the quasi "candid camera" ap- 
proach and the almost inevitable lighting difficulties, many scenes 
which are later cut together display a wide range of exposure and color 
deviations. To include all the necessary color and exposure correc- 
tions when release printing from such originals would be likely to in- 
volve too many operations to prove economical. 

This report concerns an investigation into the possibilities of in- 
cluding color and exposure compensation plus counteracting measures 
against second generation contrasts in special 16-mm key film inter- 
mediates. Where relatively few such printing intermediates are re- 
quired, it should not prove uneconomical to employ step printing, 
scene-to-scene color corrections, masking techniques and other cus- 
tom-job methods. To this end, an early model 16-mm Depue step 
printer was rebuilt to provide independent timing for each color layer 
of the duplicating stock to be printed. Exact registration facilities for 
masking operations were added. 

Various other auxiliaries were necessary. Timing viewer boxes 
were equipped with 3500 K (degrees Kelvin) light as an alternative to 

* Throughout this report, "key" films refer to intermediates made by the 
methods to be outlined. 

PRESENTED: April 8, 1949, at the SMPE Convention in New York. 


the conventional yellow light source. Color filters peaked at the 
wavelengths used in the tricolor printer were fitted to a Western 
Electric densitometer for color timing tests. A wafer-thin selenium 
cell probe, connected to a 50-microampere meter, provided the basic 
correlation of the printing exposure in each primary color. 

Before describing the tricolor printer, it is desirable to outline some 
of the corrections that it is believed a key film should possess for re- 
lease printing service. 

1. The key film should release print on one light. 

2. The general color-casts of significant images should be main- 
tained on a scene-to-scene basis. 

3. The deviations in the relative transmissions of significant hues 
which occur in both first and second generation duplicates should re- 
ceive a combined compensation in the key film. 

4. The key film should exhibit reduced chromaticity if a hue shift 
in a high chromaticity color is inevitable. 

5. A chromaticity increase should be available for scenes that were 
overexposed in the original. 

6. Underexposed and shadow areas require sufficient relative 
density reduction in the key film to expose correctly the release 

7. The general gray gamma of the key film should counteract the 
gamma gain of the release print to bring about a match with the con- 
trast of the original. 

8. The resolution of the key film images should not be materially 
lower than that of the original. 

Other complementary parts of the program call for integration 
with the black-and-white standards in routine laboratory use. The 
background information for color timing judgment was based on 
printer loop tests. Additional color printing data and the terminology 
used have been extensively treated by Miller, 1 Offenhauser, 2 Hanson 
and Richey, 3 and Yule. 4 


The design objectives of the experimental tricolor printer hinge on 
the provision of facilities to effect a maximum of the desirabilities for 
key films as outlined above. The step printer selected for modifica- 
tion provided the initial film handling and printing-gate facilities. 
The lamp-house shutter, lift cams, worm drives, master clutch and the 
sprockets were left unmodified. Extensive modifications were neces- 




sary for the handling of color masks concurrently with the original 
and the raw stock. Independently adjustable clutch-drives power 
the three take-up rolls. All sprocket guide-rollers and other film 
pressure-plates are lightly spring loaded to prevent sprocket damage 
should any of the three films become misaligned. The alternating- 
current motor drive system was changed to a molded V-belt drive for 
greater speed constancy. The original drop-board of the pin-inser- 
tion type was discarded for a newer Depue slide-bar type. 

In addition to the normal step-printing action, registration pilot 

Fig. 1. General view of printer. Top insert shows shadow-flasher with raw 
stock route, while the "original" film (center) feeds down back- to-back to the 
notched masking roll at the rear. Lower insert shows view of pressure-platen and 
projecting registration pins. The cam-operated pins shown on each side of the 
printing aperture lift the registration pins and the platen in sequence when the 
gate is shut. 


pins which span four sprocket holes were designed as part of the 
printer gate and viewer assembly. The pins were attached inside the 
viewer barrel to a spring-loaded slide-cylinder. A methyl methac- 
rylate pressure-platen, separately spring-loaded, extends from the 
slide-cylinder. The platen was made slightly convex to improve the 
printing contact of the films at the center of the frame. Cam-oper- 
ated lift-pins push the complete slide-cylinder assembly clear of the 
films during the pull-down operation (Fig. 1). The nonrigid pilot-pin 
design provides for registration between the various films which are 
separately aligned to the printer gate by an edge guide rail. 

The original lamp house was stripped and a bakelite platform 
mounted in it about an inch below the printer-gate barrel. A trian- 
gular arrangement (Figs. 2 and 3) of lamp sockets is grouped on the 
bakelite. Ductile copper brackets extend behind each socket to form 
positionable supports for three concave mirrors. The mirrors were 
turned from solid duralumin and have a 2-in. radius of curvature. 
The light sources are three 16-v, 21-cp tail-light bulbs. Spring 
holders for glass-protected Wratten filters are provided in front of each 
lamp. The light beams from the two side lamp-and-mirror combina- 
tions are redirected toward the printing aperture by two flat rear- 
surfaced mirrors mounted in adjustable holders. Completing the 
lamp-house arrangements is a countersunk door on which is mounted 
a centrifugal fan. The air blast effectively cools the Wratten filters 
and the trilamp assembly, etc., and is exhausted through the bottom 
of the lamp house and over the timing resistance banks (Fig. 3) . 

It will be apparent that no provision for optical accuracy is present 
in these lamp arrangements. A very random direction and distribu- 
tion of approximately 40 1m of tricolor light reaches the printing aper- 
ture entrance. The design postulates a light-integrating unit which 
is fitted in the printer-gate barrel to provide a uniform admixture of 
tricolor light at the film printing plane. The light-integrating unit is 
a parallelepipedon structure which is silvered externally on four sides. 
The general schematic design is shown in Fig. 2. The unit was con- 
structed from Ke- m - sheet methyl methacrylate and sealed against 
dust. The total light transmission efficiency measured 75%. When 
visually observed through the printer gate, the unit presents a series of 
shifting color-casts due to the lenticular surfaces of its diffusion plates. 
However, regardless of the degree of irregularity in color intensity 
presented at the prism end of the unit, no fringing or measurable color 
deviations appear across a test frame when film is printed. 




The arrangements for tricolor timing were designed for a minimum 
of change to the existing printing practices. The Depue drop-board 
solenoid gear was modified to drop the contact unit in steps of three at 
a time. Two additional contact brushes were added to make a triple 
contacting drop unit. The 22 riser-bars of the drop-board are con- 
nected by multiple cable and plug to three banks of adjustable tapped 
resistors located in the subcompartment on the printer (Fig. 3). 


Film Gate 

Fig. 2. 


Arrangement of the lamp-house components shown 
through the open door in Fig. 3. 

Three heavy-duty rheostats provide over-all tricolor balance for top 
light adjustments. 

The regular 22 printer lights per bar of the drop-board are divided 
into groups of seven. Each timing slide-bar can serve duty on any 
of the three colors. For convenience, sets of three consecutive slide- 
bars were assigned for the timing of red, green and blue, descending 
in that order. The timing card, as supplied to the operator, was also 
coded in triplets but no change to conventional operator practice is 




required for the setup of the slide-bars on the board. The light 
changes are initiated by regular edge notches which are applied to the 
masking and timing film roll. 

An extra triple-pole switch is located on the printer. It serves to 
connect the three lowest printer lights to the line regardless of the 
setting on the drop-board. Four of the riser bars on the board are 
not connected and are accordingly coded as zeros. This design pro- 
vides for a minimum or "base light" connection so that all lamps may 

Fig. 3. Rear view of printer with the lamp-house door open. The three 
potentiometers shown below the adjustable resistance banks provide red, green 
and blue light control with the key shown in position on the "red" potentiometer. 

470 0. K. KENDALL April 

remain lit during contact interruptions when the lights are "dropped." 
When the base light is off, it is possible, by a zero setting on the 
board, to print with the intense color of one or two of the three near- 
monochromatic lights. 

A raw stock pre-exposing unit, called a shadow-flasher, is located 
just ahead of the feed-in sprocket. It consists of a light chamber, a 
miniature lamp, a Wratten filter holder, and a film guide channel. 
The lamp is under-volted and resistance-controlled. The raw stock is 
normally routed through the shadow-flasher before joining with the 
color original and the timing film at the feed sprocket. To prevent 
the fogging of a frame when printing is stopped, the printing control 
master clutch is interlocked with a safety switch in the flasher circuit. 
A manual switch for the shadow-flasher is also provided. 


The tricolor printer furnishes additive color correction. Owing to 
the narrow pass-bands of the color filters in use, it is possible to com- 
pensate effectively for changes in color temperature of the incan- 
descent lamps when the timing resistor settings are determined. This 
timing advantage is not available with conventional single-pack color 

Each printer light change is preset to be about 70% of the next 
higher light. Seven printer lights are assigned to each of the red, green 
and blue colors. The three printer filters were selected on the basis 
of the layer sensitivity peaks of Kodachrome stock. The Wratten 
filters used are No. 29, No. 64 + No. 15, and No. 49. The top light in 
each color is adjusted by the appropriate rheostat so that tricolor 
printing of a black-and-white step-wedge is well duplicated on Koda- 
chrome. Allowance has to be made for the current processing shifts 
in dye balance. The tricolor top light exposure is normally adjusted 
to be of the order of 60 mcs (meter candle seconds), while the lowest 
light averages about 6.5 mcs (Fig. 4). 

It is essential that the normal or "center" light of a color printer be 
corrected by its color pack to duplicate an original gray scale test. 
Timing above or below the center light will shift the printed color 
balance away from " white" in accordance with the color temperature 
deviations which are produced. On the tricolor printer, this white- 
light balance is adjusted by the resistance presets to be provided 
whenever an equal number timing-light triplet is set up at the drop- 
board. For example, the nearest match to the gray scale test original 




should be printed when the board is timed red 4, green 4 and blue 4, 
i.e., the "center" light. 

The three sets of seven timing lights are coded zero to six inclusive, 
with the center light triplet adjusted to duplicate closely a test gray 
density of 1.4. The seven coequal " white" combinations provide 
30% of exposure change per step, each of which roughly corresponds 


123456 WHITE 

J. i I 1 1 J I 

65 9-5 138 20 29 42 60 MCS 



Fig. 4. Composite of relative gamma changes produced by "shadow-flash- 
ing." The printer exposures in terms of "white" timing lights are indicated 
relative to the resultant density of Kodachrome stock 5265. 

to about three standard lights on a black-and-white printer. Single- 
color light changes on the tricolor printer give rise to relative hue 
shifts of approximately 10%. This has proved to be a satisfactory 
unit of change for average color-cast corrections. 

All adjustments to the shadow-flasher unit are based on film den- 
sity measurements. Small adjustments are effected through series re- 
sistors. The initial light output of 800-mcs exposure capacity is re- 

472 O. K. KENDALL April 

duced by a pack of neutral density and color niters to about 0.6 mcs of 
equivalent cyan exposure. Less than 10% of the relative red exposure 
is transmitted. The green intensity is adjusted to provide about a 
100% increase in green transmission over the unexposed film density 
value. The blue density is lowered by about 140% increase in trans- 
mission. The determining factors are discussed later under "Pre- 
Exposure Considerations." 


It was desirable to introduce a minimum of change from the exist- 
ing black-and-white routine. A Kodak lib Sensitometer, balanced 
to expose a good gray scale, was supplied with six additional sharp- 
cutting Wratten filters. All filters were corrected by neutral density 
to a common transmission value, and all exposures were compensated 
to be equal to the gray scale exposure. 

The following Wratten filters are used: 

Wratten No. 15 Yellow 530 to 700 IHM 

64 Cyan 440 to 540 mn 

34 Magenta 440 and 680 m/z 

29 Red 680 m^ at peak 

64 + 15 Green 540 m/x 

49 Blue 440 m/* 

Relative transmission relationships are measured on the Western 
Electric densitometer fitted with tricolor filters. For convenience, all 
the hue exposures are plotted in terms of the equivalent gray gamma. 

These color-scale exposures provided informative hue transmission 
relationships for timing information. For example, blue, green and 
red gammas, when examined at the step which is equivalent to 20 
mcs of gray exposure, exhibited blue, green and red color densities of 
3%, 7J^% and 30% transmissions respectively. The lower chromatic- 
ity mid-tones, usually encountered in practice, tend to equalize such 
transmission differences. However, the familiar red build to flesh 
tone shadows is violently accentuated in a second generation print 
and the fault cannot be ignored. 

An unexpected effect was disclosed in tricolor gammas of the cyan 
test (Fig. 5) . The density to red was found to increase whenever the 
cyan exposure corresponded to an equivalent gray of 0.35 mcs or more. 
A cyan pre-exposure of this value was found largely to prevent a red 
gain in the corresponding steps of yellow gammas. Accordingly, the 
shadow-flasher pack is. prepared on the basis of the cyan and yellow 




The wafer-probe light meter is used weekly to establish tricolor 
printer exposure values co-ordinated with the current processing 
drifts through the six color gamma tests. A silver-image loop of a res- 
olution chart and a gamma strip is used to test the seven "white" 
printer-light timings. 



LOO. E. 



Fig. 5. Tricolor densities for a cyan sensitized gamma showing a virtual nega- 
tive red exposure effect starting at E. No equivalent crossover appeared in char- 
acteristic curves for other hues. 


Balanced color originals which exhibit faulty exposures are given 
almost conventional black-and-white timing practice. For timing 
purposes, any horizontal equality of the tricolor printer timing 
numbers is regarded as a single printer light, for example, blue 3, 
green 3 and red 3 become "light three, white." However, white is not 
directly estimated. Tricolor timing depends upon the green exposure 

474 O. K. KENDALL April 

which is estimated first. The red and the blue layer exposures are 
then estimated as two corrective " color tilts" given to the horizontal 
or white balance. In fact, the expression "color tilt" has come to de- 
note the timing intent as distinguished from the original, visible color 
fault. The resultant complementary hue bias which may be generated 
in the printer is seldom directly considered in rapid timing opera- 

As an example of tricolor timing, assume that the precise red and 
green proportions of the flesh tones of a given face are to be matched 
across two scenes which have been cut together. It is decided to cor- 
rect the whiter face scene by a timing card code of B4, G4 and R5. 
This plus-one-R tilt should print with a 10% red boost in all com- 
pound colors in the original provided that layer latitude for the hue 
shift is available in the part-image considered. The until ted G4-B4 
timing would be used with indoor scenes or strong blue sky highlights. 
Originals with medium blue skies will duplicate with a shift to cyan 1 
which the red boost would tend to gray. In such a case, the color 
timer will also tilt up the blue with a B4, G3 and R4 light assignment. 
Note that this is effected by dropping the red-green slope by one light. 
This is done in order to avoid the feeling of overexposure which would 
result from a B5, G4, R5 timing. 

In practice, color printing is complicated by the existence of dual 
color-cast problems. The differences in opacity to red, green and 
blue which comprise the black-level base densities, shown in Fig. 4 at 
the 2.2 to 2.6 density region, are variables which depend on the con- 
ditions occurring at the processing time. The density differences, as 
shown, may invert to any order and may range up to a displacement 
of 20% or more between the tricolor transmissions. 

During projection, this black-level color of an original is of neg- 
ligible importance. However, when the original is underexposed, 
normally mid-tone images tend to be shifted toward the 2.0 density 
region. These images partake of the hue of the current processing 
dye-balance, while images whose densities range around the 1.5 level 
may continue to exhibit the more noticeable color-cast effects of tri- 
layer mismatch in the original shooting color temperature. 

It will be apparent that whenever an underexposed sequence is 
"timed up" in printing, the prevailing process hue will modify the 
color of the restored mid-density images. At the same time, the orig- 
inal mid-tone color-cast, if any, will tend to be duplicated as an un- 
related color-cast which may appear in the lighter tones. A nicety of 


good judgment is required to achieve a satisfactory compromise in the 
tricolor corrections for a key film. It should be noted that such dual 
color-casts often present a greater problem when conventional black- 
and-white timing practices are used for release printing directly from 
the color original. 

A misleading tricolor timing problem is presented when both an 
overexposure and a color-cast occur in the original. In this case, the 
color compensation selected to offset the color-cast of the middle den- 
sities may greatly modify the pastel shades of the lowest densities pres- 
ent. The lowered printing exposure which is indicated tends to du- 
plicate the overexposed near- white areas as light grays* which may be 
noticeably tinted with the color of the offset used. When such a du- 
plicate is screened, visual adaptation tends to cancel out the tint. 
Provided that the area in question is a large one such as the sky, then 
images of mid-densities on the screen will be affected by the same 
color adaptation. An illusional color-cast is thereby created which 
has the complementary tint to the printer offset employed. To 
counteract this illusion which is, of course, the same in hue as the 
color-cast in the original, it is necessary to raise the chromaticity of the 
color correction used until direct inspection shows a hint of it in mid- 
gray image-parts of the duplicate. 

The presence of color-casts in the key film is fortunately not a major 
problem when release printing. Any reel-length over-all color-cast 
may be easily offset by an appropriate color-correcting filter. The key 
film dye-balance color requires no additional color pack changes pro- 
vided the printing can be done on center light. 

Owing to the low latitudes involved, the center light duplication of 
spectral densities of the order of 2.0 tend to print below the black- 
level. The shaded print-through gamma curve* E shown in Fig. 4 in- 
dicates the average final latitude available. It will be appreciated 
that little opportunity exists for any timing exposure excursions un- 
less some method of compression is introduced. 


The flashing of the raw stock by a white-light exposure of about 
0.35 mcs will lower the tricolor gamma. As shown by curves A in 
Fig. 4, exposures below this level are not visibly recorded. The ad- 

* See Fig. 4. A completely overexposed original with a transmission of the 
order of 60% (a density of 0.2) is reproduced by light No. 1 as a gray of only 20% 
(a density of 0.7). 

476 0. K. KENDALL April 

dition of the above flash to each step not only shifts the point of 
shoulder inflection but it also produces a chromaticity limit to shadow 
details. In addition, any processing dye-balance color-cast tends to 
become more predictable a useful feature when black-and-white 
timed duplicates are directly printed from the original. Flashed raw 
stock may be used for second generation duplications but the release 
film blacks should project as visually normal. Release film shadow- 
flashing is, therefore, definitely limited to the inflection point of the 

The differences between the relative gammas of the primary colors 
has been fully discussed in the JOURNAL. 1-3 They are the cause of the 
familiar red build-up which sometimes characterizes the shadow sides 
of faces. Various approaches to the problem, such as the intentional 
printing of a cyan color-cast or the use of a red light mask, 1 have been 
employed. However, shadow-flashing with cyan light proved a con- 
venient control measure. The density to red at the black-level has 
been shown to increase when a cyan exposure is used (Fig. 5 at E). 
This process fault is most useful because of the concurrent reduction 
to the chromaticity of dark reds provided by the cyan transmission 

In practice, a higher shadow-flash than 0.35 mcs is needed for the 
key film stock. The compensation for a 2.25 red gamma which is the 
product of two generations of color printing would require a key film 
red gamma of approximately 0.45. Shadow-flashing with a pre- 
dominantly cyan light (Fig. 4, C) of about 0.6 mcs is seen to produce a 
region of gamma of this order as manifested by the curves at B. 
The release printer light is adjusted to duplicate as neutral black the 
key film shadow-flash density levels R s , B s and G 8 . The relatively 
greater density to red tR s > to G 8 B S ) of about 0.1 is required in order to 
compensate for the low cyan gamma of the final duplicate. 

The existence of the shadow-flash "floor" permits the timer to con- 
centrate on the maintenance of highlight details. Only shots possess- 
ing no near-whites require higher than a center-light exposure. The 
effect of shadow-flashing extends even to the mid-densities in counter- 
acting possible dye-casts so that the hue substitutions of the original 
become the color timer's major consideration. 

Precautions must be taken, however, not to make an automatic 
practice of undertiming because of the resolution losses which may be 
incurred. The resolution of details is not only a function of the grain 
size possessed by the silver images before bleaching; it is also de- 


pendent upon the degree of density excursion forming the image de- 
tails. For colored images, each increase in color density can be ex- 
pected to exhibit lowered resolution. The density differences upon 
which picture details depend are thus reduced by lower timing. The 
following resolution figures give some idea of the definitions to be ob- 
tained when printing pure hues through a resolution chart two lights 
below normal : 

B G R lines/mm 

Black-and-white resolution on: Center Light 4 4 4 * 55 

Red " 2 -* 17 

Green " 2 -* 25 

Blue " 2 -* 35 

A gain of up to 15 lines/mm occurred when equivalent center-light 
pure hues printed the resolution charts. The reduction in maximum 
contrast due to flashing affected the practical color resolution by a 
loss of about 10 lines. In any event, the resolution losses from shadow- 
flashing do not begin to compare with the total loss of detail that oc- 
curs when portions of an image are cut off by printing below the 
gamma shoulder of non-shadow-flashed Kodachrome film. 


While the various relative-brightness silver masks produce contrast 
reduction as outlined, 1 their routine use for one-light printing seems 
contraindicated. Such masks, either unsharp 4 or the more convenient 
sharp masks defocussed by double film base thicknesses, have printed 
uncommercial amounts of halo around those image fimbrillations 
which have extreme contrast.* Very lightly printed blue-sensitive 
masks for the reel-length equalization of skies show possibilities but 
timing complications ensue if full masking with tinted base panchro- 
matic stock is used. For obvious reasons, negative masking cannot 
be used with overexposed scenes. Also the whole feasibility of mask- 
ing is modified by the additional corrections which are required for 
the second generation duplicate. 

A different proposition altogether is the use of color film for the 
masking of overexposed scenes. Wherever such scenes occur, it is 
desirable to accentuate the relative hues and contrasts. A conven- 
tional color print cannot be used as a mask, however, as normally ex- 

* In this connection, it has been noted that, for masking registration purposes, 
neither the film direction nor the perforation sides are interchangeable. 

478 0. K. KENDALL April 

posed images are always present in overexposed shots. For this 
reason, it is important to limit both the chromaticity and the gamma 
of the positive mask. This may be achieved by preflashing the mask- 
ing stock. A white flash of approximately 3 mcs (see step D, Fig. 4) 
will set up a maximum minus density of about 10% transmission to all 
hues. The optimal value is not critical, nor is the time lapse between 
flashing and use, but the fogged stock should be pretested to ensure a 
neutral gray. 

The overexposed original is printed onto this stock by center-light 
' or lower. Provided that exact registration is used, the resulting pastel 
colored mask is an effective compensation means. The low mask 
gamma (about 0.35 7, Fig. 4, curve E) largely avoids the problem of 
halos. Also the relative color and density shifts co-operate visually. 
The chroma-boosting is such that no exception to the routine timing 
and shadow-flash procedure is required. Full gamma Kodachrome 
masks used in conjunction with extreme printer-light tilts have pro- 
vided endless combinations for special effects. 


Expressed in printing vernacular, the production of a 16-mm key 
film intermediate follows this pattern: 

The original is checked and very thin scenes are papered. Using 
the white fogged stock (3 mcs), the chroma-boosting masks are center- 
light printed between the papers. 

After processing, the masks are frame synchronized with their origi- 
nals by clear leader. 

A first run-through is made over the (color) light box to judge the 
printer lights and the green exposure is set up on the timing card. 
Double timing from previous notches on the original is avoided by 
notching on the masking roll. 

The color corrections are next estimated and the red and the blue 
timing tilts are entered on the timing card. Overexposed scenes are 
superimposed on their masks when viewed for timing. 

The notched mask is then synchronized tail-out behind the original 
in the tricolor printer. Finally, the tricolor drop-board and the 
shadow-flasher are preset, the raw stock is threaded and the key film 
is printed. 


The major disadvantage to the practice of printing from key films 
lies in the resolution losses incurred at the 2.0 density region. A 


limitation to the maximum chromaticity also occurs at the same re- 
gion, but specific hue-shifts are considered to be less objectionable 
when the chromaticity is thus reduced. 

Other observations on test prints made from key films have demon- 
strated : 

A contrast of the same order as the original. 

Improved maintenance of original shadow image densities. 

Increased average transmission, jointly with improved reproduction 
of highlight areas. 

Reduced color-shifts on a scene-to-scene basis, when compared with 
the original. 

Reduced color-casts. 

Reduced exposure changes. 

Reduced effects from relative-brightness color-shifts. 

Increased chromaticity and contrast from overexposed originals. 

Increased chromaticity; reduced gray contrast for special effects. 

In view of the opportunities for greater production control which 
are presented by the use of an intermediate stage, it is felt that the 
key film method can be the basis of improved release quality in 16- 
mm color. 


(1) T. H. Miller, "Masking: A technique for improving the quality of color 
reproductions," Jour. SMPE, vol. 52, pp. 133-155; February, 1949. 

(2) W. H. Offenhauser, Jr., "Some notes on the duplication of 16-mm integral 
tripack color films," Jour. SMPE, vol. 45, pp. 113-134; August, 1945. 

(3) W. T. Hanson, Jr., and F. A. Richey, "Three-color subtractive photog- 
raphy," Jour. SMPE, vol. 52, pp. 119-132; February, 1949. 

(4) J. A. C. Yule, "Unsharp masks," /. Phot. Soc. Amer., vol. 11, pp. 123-132; 
March, 1945. 


MR. WILLIAM OFFENHAUSER: I think this paper is a great contribution. I 
don't think at the moment that its import is fully appreciated, but we have for 
many years needed some analysis of second and later generation Kodachrome 
prints and some means to overcome their difficulties. There are a number of 
very fortunate things on the horizon. Now that the color control problem is 
reasonably well within view of solution, we have left the major problem of de- 
terioration in resolving power; and from all that I have been able to hear, mostly 
through what we might call the underground, that solution, too, is in sight. 
So it may not be very many years before we will be able to accomplish the objec- 
tive of having an original, storing it away for many years and yet being able to 
derive, let us say, third generation and further generation prints for use in the 
16-mm fields. The goal is in sight, 

Illuminating System and 
Light Control for 16- Mm 
Continuous Optical Printer 


Summary A continuous optical printer for producing highly corrected re- 
lease prints at rapid printing speeds from 16-mm color originals is being de- 
veloped and tested by the Eastman Kodak Co., Rochester, N.Y. All informa- 
tion on level and color balance of the printing light for each scene may be 
stored on the master film and the cuing device eliminates the necessity of 
notching the film. Also incorporated is a sprocket which automatically ac- 
commodates highly shrunk and damaged film inter-cut with fresh originals. 

THE INCREASING DEMAND for high-quality release prints from 16-mm 
color originals poses problems not readily solved by modifying 
existing printers. Since some original camera film may be included 
in the printing master from which the release prints are to be made, the 
requirements imposed on a color printer are indeed stringent. 

First, release prints must be turned out at high speed; yet the ex- 
tremely valuable original film must not be worn out until a profitable 
number of copies has been obtained. This requirement has been met 
in this experimental model by using precision gears to register the 
two sprockets of a continuous optical printer. Thus the film can be 
driven at high speed without the characteristic wear and damage of 
an intermittent film drive, or the abrasion of contact printing. The 
handling of the original film is further facilitated by an invaluable 
device which we call our "compensating sprocket." This is a sprocket 
which changes pitch continuously throughout the angle of film "wrap" 
and hence smoothly handles a wide range of shrinkages as well as 
worn or broken perforations. 

Secondly, it has been found empirically that any change in ex- 
posure or color balance must be accomplished within one frame in 
order not to be objectionable when the print is projected. If, for 
example, the speed of the printer is set at 100 feet/min, then 
changes in color balance or exposure must be accomplished within 
15 milliseconds; higher printing speeds obviously would require a 
proportionately faster action. Since the original film may vary 
PRESENTED: October 11, 1949, at the SMPE Convention in Hollywood. 



widely in exposure and color balance because of the unavoidable 
conditions under which the shots are made, a wide range of correction 
is needed in each color in a release printer. But small corrective steps 
in each color are required in order that the continuity of color from 
scene to scene be accurately maintained. Thus, the total number of 
possible combinations of printing level and color corrections is 

If one chooses values of optical density according to the geometrical 
progression 1, 2, 4, 8, 16, up to n and arranges to interpose any com- 
bination of these densities into a beam of light, then he has made 
available 2 n equal steps of light intensity. If then three beams of the 
same intensity, but each of a different primary hue, are so filtered 
and combined at the slit of the printer, a unit density may be chosen 
to give the desired fineness of color correction, and the value of n 
to give the necessary range. 

In our experimental model, three condenser relay systems with 
the aid of multiple-layer selective reflectors combine three primary 






Figure 1. 


beams, while solenoids move the corrective filters with the required 
speed (see Fig. 1). In order to use this range and speed of control, 
one must first assign to each scene of the original film a printing level 
specified in terms of the desired intensity of each of the three primary 
beams. This information must then be stored in such a way that it 
can reappear and alter the printing light at the beginning of each 
scene. Again selecting 100 feet/min for purposes of example, a 
one-foot scene would allow six-tenths of a second for "clearing the 
memory " and feeding in the prepared information of the following 

The memory system selected for storing the scene-by-scene printing 
information is a narrow track of magnetic material applied to the 
master film between the perforations and the edge of the film. If 
five filters are used in each of the three primary beams as described 
above, giving 32 steps in each color, then the conditions for printing 
any given scene may be specified by demanding that certain of the 
15 filters (5 in each color) be interposed in their primary beams. Thus 
is established not only color balance but exposure as well. In a special 
viewer-recorder equipped with a magnetic recording head, we put 
down on the magnetic track at the beginning of each scene a series of 
15 pulses corresponding to the 15 filters each positive pulse calling 
for that particular filter to appear in its colored beam for the next 
scene, and each negative pulse specifying that the corresponding 
filter remain out of its beam. Positive and negative pulses are pro- 
duced by opposite polarity on the coil of the recording head. Fol- 
lowing this sequence of pulses, a triggering pulse is put down on the 
magnetic track, the function of which will become apparent below. 

As the "color-timed" master film is run through the printer, the 
fifteen pulses pass a magnetic playback head on the printer before the 
corresponding scene reaches the printing "gate." This pattern of 
positive and negative pulses is electronically stored until the triggering 
pulse reaches the playback head, at which time the pattern is de- 
livered to the solenoids that actuate the filters, accomplishing the en- 
tire light change within the first frame of the new scene. 

By this approach to the problem not only is the printing informa- 
tion permanently stored as an integral part of the original film, but 
also the cuing of light changes is accomplished without notching the 
film. Both the printing information and the triggering pulse may be 
erased and re-recorded as desired, facilitating subsequent editing 
and refinement of printing conditions to give a maximum of color 
continuity throughout the release print. 

New Brenkert Projection System 
For Drive- In Theaters 





Summary Some of the limitations and handicaps in illuminating large 
drive-in theater screens are discussed. The basic requirements for ade- 
quate screen lighting are reviewed and a newly developed arc light and pro- 
jection system meeting those requirements are described. 

during the war and soared to undreamed-of heights. Perhaps 
one reason for this rapid rise in popularity was that the drive-in 
theater gave people, who were in a war plant all day, a chance to relax 
and see a picture out in the open air and in a picnic-like atmosphere. 
Since the war, the popularity of the drive-in theater became even 
greater and, as a result, there is still a wild scramble to build drive-in 
theaters in all parts of the country. 

Vast improvements have been made in the design of drive-in 
theaters compared to those built prior to the war: Buildings and 
grounds have been beautified; services such as ultra-modern snack 
bars and kiddies' playgrounds have been incorporated as added at- 
tractions; and parking areas have been increased in size so that many 
theaters will now accommodate over 1,000 cars. RCA pioneered in 
designing and building sound reproducing equipment especially for 
drive-in theaters so that good quality sound could be reproduced in 
every car through a neat and attractive in-car speaker. Producing a 
picture that can be clearly and easily seen by the occupants of all the 
cars in drive-in theaters, however, has always beeen a difficult task up 
to the present time. 

There are a number of factors that contributed to the difficulty in- 
volved in high-quality projection for large drive-in theaters: 

1. Long viewing distances between the cars in the rear ramps and 
the screen, in some cases over 850 ft. Detail of the picture, even 

PRESENTED: October 14, 1949, at the SMPE Convention in Hollywood. 



though it may be 70 ft wide, is rapidly lost at these distances because 
visibility is lessened and the elements also interfere and cut down on 

2. Long projection throws of 300 ft or more. Here again interfer- 
ence of dust, mist, fog and rain have a deleterious effect on picture 

3. Extremely large picture areas, some over 3,000 sq ft. It is ex- 
tremely difficult to provide adequate illumination over picture areas of 
this size so that the people in cars in the rear and on the sides of the 
parking area can see details. 

4. Wide viewing angles, particularly at the ends of the front ramps. 
This is a limitation only for those cars at the extreme sides and reason- 
ably close up front, and is a limitation that cannot be remedied ex- 
cept by changing the shape of the parking area with a considerable 
loss in productive area. 

5. Moonlight and other interferences such as electric signs, highway 
lights and street light. 

The net result of these limitations is that most drive-in theaters 
have pictures that are considerably inferior to those in regular indoor 
theaters. It is also quite apparent from the nature of these limita- 
tions that by providing a sufficient amount of projected light the 
quality of the drive-in theater picture could be considerably enhanced. 
Very little work has been done in establishing the lighting values for 
outdoor screens; consequently, there are no standards such as we 
have for indoor screens. It is generally conceded that for the same 
viewing distances, large screens do not need so much illumination as 
smaller screens to permit the viewer to see an equivalent amount of 
detail. Recent observations and measurements of present screen il- 
lumination in outdoor theaters indicate that a 50% to 100% increase 
from present levels is necessary if the picture quality is to be mate- 
rially increased. Such an increase in illumination will bring the light 
on large screens up to 6 ft-c at the center; this will provide reasonably 
satisfactory quality of projection under all but the most extreme con- 
ditions. A field flatness of at least 70% is desirable, but reasonably 
good picture quality can be had with a field flatness as low as 60%. 
Lighting values such as these can be obtained on screens 35 to 40 ft in 
width using standard projector mechanisms and suprex-type arc 

A super high-intensity arc lamp, burning 13.6-mm high-intensity 
positive carbons at 150 amp operated in combination with a projector 


equipped with double disc-type shutters and using //2.0 condensers 
and projection lenses, will deliver approximately 7.6 ft-c at the center 
of a screen 40-ft wide, with an 80% light distribution. 1 This same 
projection system will deliver approximately 3.5 ft-c at the center of 
a 60-ft screen with the same light distribution. Increasing this light 
intensity beyond these values is not very practical mainly due to heat 
problems that arise and the probable damage to film, unless some ade- 
quate means of cooling is provided. 

Heat-absorbing glass is sometimes used to reduce the heat on the 
film, but all known glass heat filters absorb a considerable amount of 
the visible light rays as well as change the color of the projected light. 
The net result is a visible loss of considerably more than that indi- 
cated by a photometer with a Viscor filter. Light measurements 
made using arc lamps operated at 170 amp with heat filters actually 
give screen results which are inferior to those when using the same 
arc lamp operated at 150 amp without the heat filter. Other methods 
of cooling, such as by water-cooled apertures, do not provide adequate 
protection to the film. 

The light projection problem therefore resolved itself into the fol- 
lowing objectives: 

1. Obtaining a carbon capable of producing approximately 26,000 
lumens on the screen without the projector running when operated at 
its rated current so as to obtain at least 6 ft-c at the center of a 60-ft 

2. Designing an arc lamp capable of operating satisfactorily using 
high-current carbons of this type,*yet flexible enough so that it will 
perform equally well when operated at lower currents. 

3. Designing a new film trap and gate assembly with associated 
cooling system permitting transmission of maximum light from the arc 
lamp to the screen. 

All of these objectives have been met with the introduction of new 
projection and arc lamp equipment known as the Brenkert Super- 
tensity system and developed by the Brenkert Light Projection Com- 
pany. Incorporated in this system are methods for adequate cool- 
ing of the arc lamp, projector and film so that this system can be 
operated with high amperage carbons at the full current rating of the 
carbon and without damage to the film or the equipment. Using this 
system with the 13.6-mm super high-intensity carbon it was possible 
to use the maximum amount of light developed when these carbons 
were operated at their maximum rating of 170 amp without employ- 

Fig. 1. Supertensity Lamp. 

Fig. 2. Interior of Supertensity Lamp; showing positic 
of lenses, carbon feeds, mechanisms and air outlets. 




ing light-wasting and color-changing glass heat filters. This was 
more light than was ever used before without glass heat filters to pre- 
vent film damage. 

With this new Brenkert Supertensity system, however, even more 
light could be used, without damaging film, than could be developed 
by the 170-amp 13.6-mm super high-intensity carbon. The new 13.6- 
mm "Hitex" Super carbon recently announced by the National Car- 
bon Division offers such a possibility of increased light. This carbon 
is rated at 170 to 180 amp; the trim is complete, using a J^-in. heavy- 
duty Orotip negative. When operated at 180 amp in the Supertensity 
system approximately 7 ft-c can be projected on the center of a 60- 
ft screen with approximately 70% screen distribution. The light 
produced when operating this new "Hitex" carbon at 180 amp is much 
whiter than that produced by the Supertensity system and standard 
super high-intensity carbon when operated at 170 amp, so that to the 
eye the light on the screen appears greater than indicated by a photom- 


In designing the lamp house consideration was given to such things 
as ease of operation, cooling, placement of the carbon feed mechanism 
and appearance. The housing is extremely large by comparison to 
other arc lamps, having a content of approximately 10 cu ft, which 
aids in cooling and at the same time greatly facilitates operation, serv- 
icing and cleaning. 

The walls of the housing are hollow, with intake ventilation vents 
around the bottom through which air enters, passes up through the 
walls and exhausts through vents located around the stack port at the 
top. Forced ventilation is supplied to the interior of the lamp by a 
fan which is driven by the carbon feed motor. Air from this source is 
passed through ducts which are part of the base casting and terminate 
under the positive feed, the carbon feed and the negative carbon 
holder. The entire lamp house is constructed of solid aluminum cast- 
ings, completely lined and ventilated to prevent its becoming exces- 
sively hot and warping out of shape. This is especially important to 
permit an accurate arc image to be projected on the arc-image screen. 

The positive carbon feed mechanism uses a three-roller head which, 
while rotating the carbon, feeds it toward the arc ; grease-packed ball 
bearings are used throughout. Many new and very desirable fea- 
tures are incorporated in the design of this new positive carbon feed 

Fig. 3. Rear view of Super-tensity Lamp; showing arc current meter, 
hand carbon feeds and motor feed mechanism. 


Fig. 4. View of BX-80 Projector; showing air jet position. 


mechanism. The three-feed rollers are % in. in length and the full 
length of each roller contacts the carbon at points equidistant on the 
circumference of the carbon with a pressure of approximately 50 Ib 
existing between the rollers and the carbon. Thus, positive feeding 
of the carbon is assured at all times; tests at the factory have proved 
that it is impossible to stall the carbon while the feed mechanism is in 
operation. Accurate alignment of the carbon is assured regardless 
of the length of trim because of the three-suspension method of hold- 
ing the carbon and because of the proximity of the rollers to the arc. 
The positive carbon brushes are located directly in front of the feed 
rollers. The carbon passes through a high heat baffle to the crater 
point. This baffle protects the mechanism and serves as an air guide 
to prevent the forced air ventilation from disturbing the arc. The 
linear dimensions of the feed mechanism brushes and baffle are short, 
permitting the carbon to be burned to a stub of approximately 3J^ 
in. Current feeds to the carbons are symmetrically arranged to as- 
sure perfect magnetic balance regardless of operating current values 
and without the use of auxiliary permanent magnets. This results in 
a white and unwavering light on the screen at all operating current 

The negative carbon feed is placed alongside the positive feed. An 
L-shaped arm supports the negative carbon clamp. This construc- 
tion keeps all gears and moving parts away from the intense heat of 
the light beam, reducing heat deterioration of the parts considerably. 
The forced air cooling makes it possible to eliminate the necessity of 
using graphite lubricant; ordinary motor oil is used throughout the 
lamp for lubrication. 

The carbon feed motor, the ventilating fan and all controls are 
placed at the rear of the lamp with the hand-feed cranks extending 
through the cover. The arc current meter is also visible through the 
cover, permitting easy reading of the arc current at all times. The 
carbon positions are indicated on the large ground glass screen which 
is located on top of the lamp, and can easily be seen at all times from 
any position in the booth. 

Manual striking of the arc is accomplished by operating a lever lo- 
cated at the right rear of the lamp. This operation raises the nega- 
tive to contact the positive and then allows it to fall back quickly into 
operating position when the lever is released. Using this type of arc 
striker, the arc can be struck at full rated operating current without 
danger of damaging or cracking the arc crater. 


The lamp is equipped with high-speed //2.0 condensers. The con- 
denser next to the arc is fused quartz; the front one is Pyrex. This 
complete lens assembly is adjustable laterally, vertically and forward 
and backward so that it can be properly positioned to obtain maxi- 
mum screen brightness and uniform distribution. The condensers 
are protected during striking by a hand-operated dowser. This type 
of condenser system is recommended wherever maximum light on the 
screen is the prime factor. The standard, less expensive Brenkert 
condensers can also be used in this lamp for indoor theater use where 
flatness of field is the prime factor and high efficiency light trans- 
mission is not so important. 


Safely transmitting the unrestricted light from the arc lamp through 
the projector and film necessitated some revolutionary changes in de- 
sign of the film side of the projector. In addition to protecting the 
film from heat damage, the parts of the projector exposed to the in- 
tense heat in the light beam had to be constructed of very high heat- 
resisting materials and had to be adequately cooled. The projector 
selected for this modification was the Brenkert De-Luxe BX-80. In 
addition to its inherent rugged construction, the BX-80 is equipped 
with double rear disc-type shutters. This type of shutter construc- 
tion has the advantage of passing more light with less heat on the 
film than a projector mechanism using front and rear disc-type shut- 
ters or barrel-type shutters. 

The film trap and gate assembly is designed and constructed to cope 
with the heat problem by making use of high-heat-resistant reflector 
baffles at the point of light entry to the film trap assembly. Fire 
shutters are constructed of metal which is highly heat resistant. 

Cooling of the film and gate assembly is accomplished by using 
compressed air. The effectiveness of this type of cooling is dependent 
on several factors such as air pressure, velocity and the manner in 
which the air is directed onto the film. The entire film trap assembly 
is also effectively cooled by this air. Water-cooling the metal parts 
of the film trap around the aperture was tried but was discarded as 

The maximum heat in the light beam at the aperture is on the emul- 
sion side of the film. This heat, which is energy absorbed, is propor- 
tional to the density of the image on the film. It is also somewhat 


greater at the center of the aperture than at the sides. Distribution 
of the air stream must, therefore, be such that sufficient air contacts all 
parts of the film so as to absorb the heat and carry it away. The 
metal parts of the film trap and gate are automatically cooled by the 
air as it is exhausted from around the film trap assembly. 

The proper application and distribution of the air stream is ob- 
tained by specially designed jets. They are placed front and rear of 
the film and positioned so that air is ejected horizontally against the 
film. The jets are on the side of the gate next to the center plate of 
the projector and the air is blown toward the operating side of the 

The volume of air directed against the emulsion side of the film is 
somewhat greater than against the front or base side of the film. 
Flutter and consequent focus trouble are completely eliminated by 
proper construction and placement of these jet nozzles. 2 The angle 
at which the air strikes the film is also important in this respect. 

Air is piped to the projectors from a pump which is usually in- 
stalled in the motor generator room. One reason for using an air 
pump is to supply air completely free of oil, something which cannot 
always be done with a regular piston-type compressor. A silencer 
and cleaner are attached to the air pump to reduce the noise and dust 
in the air. 

A number of these installations have been made in several parts of 
the United States, many of them more than one year ago. The 
picture quality at the drive-in theaters using this system is far su- 
perior to anything ever seen before in a drive-in theater. 

Lighting standards which were mentioned previously in this paper 
are not intended to be optimum, but it has been proven that, if the 
larger drive-in theaters would bring their screen illumination up to 
these values, a very desirable improvement in projection could be 


(1) National Projector Carbons, 4th Ed., National Carbon Div., Cleveland, 
Ohio, 1949. 

(2) F. J. Kolb, Jr., "Air cooling of motion picture film for higher screen illumi- 
nation," Jour. SMPE, vol. 53, pp. 635-664; December, 1949. 

Note on Metol Analysis 
In Photographic Developers 



THE METHOD of Brunner, Means and Zappert 1 for the determina- 
tion of metol in Ansco Color Positive first developer, A-502, was 
found by us to give a titration curve with no -true inflection point. 
Instead, there was a region of about 0.5 ml in which the inflection 
might occur; therefore, the error in the determination may be quite 
large. By substituting acetic acid for water as solvent for the titra- 
tion, and 0.1 N sulfuric or perchloric acid in acetic acid for the con- 
ventional 0.1 N hydrochloric acid, a better inflection point may be 
obtained. No changes except those already mentioned were made 
in the procedure given by Brunner, Means and Zappert. 

Briefly, the theory is that whether a substance is an acid or base 
depends on the solvent in which it is dissolved. 2 - 3 Nitric acid is 
commonly regarded as a strong acid, but when nitric acid is dissolved 
in concentrated sulfuric acid it acts as a moderately strong base. 

Metol is an amino phenol. In water the amino group is a weak 
base and the phenol group is a very weak acid. The net effect is a 
weakly basic reaction towards acids. If, however, acetic acid is used 
as the solvent, the acidity of the phenol group is completely masked 
while the basicity of the amino group is enhanced. 

The titrant must be a stronger acid than acetic acid ; sulfuric and 
perchloric acids are very convenient and a 0.1 N solution of either in 
glacial acetic can be accurately standardized against diphenylguani- 
dine potentiometrically or with methyl violet indicator. 

A comparison of two titrations, one in water and the other in acetic 
acid, is shown in the accompanying graph. On calculating the change 
in emf per milliliter at the equivalence point, it was found that for 
water the ratio was 64 mv per ml and for acetic acid, the ratio was 112 
mv per ml. Furthermore, these values applied over a region of 0.5 
and 0.2 ml respectively. It can be seen that the end point can be 
found more closely with acetic acid as solvent. 

A CONTRIBUTION: Submitted December 28, 1949. 




It should be noted that a precipitate forms when the titration with 
eerie sulfate is performed. The end point is not as sharp as in water, 
and either a separate sample should be prepared or an aliquot part of 
the combined extracts of metol and hydroquinone may be used. 






Comparison of typical titration curves from a 
metol-water solution and a metol-acetic acid solution. 


(1) A. H. Brunner, Jr., P. B. Means, Jr., and R. H. Zappert, "Analysis of 
developers and bleach for Ansco Color Film," Jour. SMPE, vol. 53, pp. 25-35; 
July, 1949. 

(2) L. P. Hammett, Physical Organic Chemistry, Chap. II, McGraw-Hill, 
New York, 1940. 

(3) N. F. Hall and J. B. Conant, "A study of superacid solutions: I," 
J. Amer. Chem. Soc., vol. 49, p. 3047; 1927. 

New American Standards 

Six NEW American standards, approved by the American 
Standards Association on March 14, 1950, appear on the 
following pages. The four which deal with 16- and 8-mm 
camera and projector apertures were published as proposed 
standards in the March, 1949, JOURNAL, for a period of trial 
and comment. No criticism of the proposals was received; 
therefore no change in the technical content has been made. 

The Standard for Mounting Frames for Theater Screens 
(Z22.78) was developed by a Subcommittee of ASA Sectional 
Committee Z22, and is being published here for the first time. 
The need for a standard of this type became apparent in 1946 
when the revision of Standard Dimensions for Theater Screens 
Z22.29 was undertaken. The new standard describes good 
current practice and will aid manufacturers and theater owners 
in selecting the appropriate frame for any particular appli- 

The Standard for 16-Mm Sound Projector Test Film (Z22.79) 
is also being published for the first time. It was developed by 
the joint Test Film Committee of the Motion Picture Research 
Council and the Society as a revision of War Standard Z52.2. 
It describes a 16-mm version of the 35-Mm Theater Sound 
Test Film, familiar to many members as the old " Academy " 
test reel. The primary difference between this American 
Standard and the old War Standard is of an editorial nature. 
The detailed procedure for selecting appropriate sound test 
samples is now covered in the American Standard for the 
35-Mm Film Z22.60 which was approved in 1948 and was 
published in the November, 1948, JOURNAL. 

One other important change concerns the re-recording char- 
acteristic to be used in making up the 16-mm film. During 
the war there was no agreement as to what high-frequency 
equalization should be used in the 16-mm re-recording channel. 
Now, however, the major studios have reached an agreement, 
and the recommendations have been published as the Research 
Council Bulletin N-l.l. 



American Standard 

Location and Size of Picture Aperture of 
16-Millimeter Motion Picture Cameras 


Revision of 
Z2 2.7- 1941 


UDC 778.53 

Page 1 of 3 pages 

This standard applies to both silent and sound 16-mm. motion picture 
cameras. It covers the size and shape of the picture aperture and the rela- 
tive positions of the aperture, the optical axis, the edge guide, and the film 
registration device. The notes are a part of this standard. 









A (measured perpen- 

dicular to edge of 


0.201 minimum 

5.1 1 minimum 


B (measured parallel 

+ 0.006 

, ..,+0.18 

to edge of film) 

- 292 - o!oo2 

/>4 -0.05 



0.31 4 0.002 

7.98 0.05 



0.1 10 minimum 

2.79 minimum 



0.1 25 0.002 

3.1 8 0.05 



0.1 75 0.002 

4.44 0.05 



0.474 0.002 

1 2.04 0.05 



0.773 0.002 

19.63 0.05 



1.072 0.001 

27.23 0.03 



0.020 maximum 

0.51 maximum 


Approved March 14, 1950 by the American Standards Association Incorporated 
Sponsor: Society of Motion Picture and Television Engineers Incorporated *Un 

I Drttm.l CUi.ificili.rt 

Copyright 1950, by American Standard Assn., Inc., reprinted by permission of the copyright holdtr. 


American Standard 

Location and Size of Picture Aperture of 
16-Millimeter Motion Picture Cameras 

Rrg. V. 5. fat. Of. 


Revision of 
Z2 2.7 -1941 


UDC 778.53 

Page 2 of 3 pages 

The angle between the vertical edges of the aperture and the edges of 

normally positioned film shall be degrees, Vz degree. 

The angle between the horizontal edges of the aperture and the edges of 

normally positioned film shall be 90 degrees, Vz degree. 

Note 1 : Dimensions A, B, and R apply to the size of the image at the plane 
of the emulsion; the actual picture aperture has to be slightly smaller. The 
exact amount of this difference depends on the lens used and on the sep- 
aration (dimension G) of the emulsion and the physical aperture. G should 
be no larger than is necessary to preclude scratching of the film. The 
greatest difference between the image size and aperture size occurs with 
short focal-length, large diameter lenses. 

Dimensions A and B are consistent with the size of the images on a 
1 6-mm. reduction print made from a 35-mm. negative with the standard 
2.15 reduction ratio. 

It is desirable to hold the vertical height of the actual aperture to a 
value that will insure a real (unexposed) frameline. This results in less 
distraction when the frameline is projected on the screen than is the case 
when adjacent frames overlap. 

Note 2: The edge guide is shown on the sound-track edge. This location for 
it has the advantage that the rails bearing on the face of the film along 
this edge and also between the sound track and picture area can be of 
adequate width. Disadvantages of this location for the edge guide are 
that, because film shrinkage and tolerances affect the lateral position of 
the perforations, the pulldown tooth must be comparatively narrow and 
will not always be centered in the perforation. 

The guide can be on the other edge, adjacent to the perforated edge 
of sound film. With the guide at this edge, the width of the pulldown tooth 
does not have to be decreased to allow for shrinkage. However, because 
of variations introduced by shrinkage of film, this location for the edge 
guide has the important disadvantage that it makes extremely difficult 
the provision of rails of adequate width to support the sound-track edge 
without encroaching on, and consequently scratching, the picture or 
sound-track area. (See Section 3, Proposals for 16-mm. and 8-mm. 
Sprocket Standards, Vol. 48, No. 6, June 1 947, Journal of the Society of 
Motion Picture Engineers). 


American Standard 

Location and Size of Picture Aperture of 
16-Millimeter Motion Picture Cameras 

It.c. V. S. Pal. Of. 


Revision of 


ZM. 13- 1941 
UDC 778.53 

Page 3 of 3 pages 

The film may be pressed against the fixed edge guide by a spring, by 
the tendency of the film to tilt in.the gate, or by other means. In the second 
case, there is a fixed guide for each edge of the film. The important point 
is to have the film centered laterally on the optical axis. 

Dimension C is made slightly less than half the width of unshrunk film 
so that the film will be laterally centered if it has a slight shrinkage at the 
time it is run in the camera. This is the normal condition. As indicated by 
the above discussion, C may be measured in either direction from the 
vertical centerline. 

Note 3: Dimension F must be maintained only when a photographic sound 
record is to be made on the film that passes through the camera; other- 
wise F may be disregarded. 

Note 4: The K dimensions are measured along the path of the film from the 
horizontal centerline of the aperture to the stopping position of the regis- 
tration device. Both the dimensions and tolerances were computed to 
keep the frameline within 0.002 to 0.005 inch of the centered position for 
films having shrinkages of 0.0 to 0.5 per cent at the time they are exposed 
in the camera. For any given camera, use the value of K corresponding to 
the location of the registration device. 

If the film does not stop exactly where the film registration device leaves 
it, because of coasting or some other cause, a slight adjustment of the 
value of K will be necessary. This will be indicated if film that has a shrink- 
age of 0.2 to 0.3 per cent when it is run in the camera does not show a 
properly centered frameline. From such a test, the amount and direction 
of the adjustment can be determined. 

Note 5: "Optical axis of camera" is defined as the mechanical axis or cen- 
terline of the sleeve or other device for holding the picture-taking lens. 
Except for manufacturing tolerances, it coincides with the optical axis 
of the lens. 




American Standard 

Location and Size of Picture Aperture of 
16-Millimeter Motion Picture Projectors 

.<!. V. S. Pal. Og. 


Revision of 


UDC 778.55 

Page 1 of 3 pages 

This standard applies to both silent and sound 16-mm. motion picture 
projectors. It covers the size and shape of the picture aperture and the rela- 
tive positions of the aperture, the optical axis, the edge guide, and the film 
registration device. The notes are a part of this standard. 












A (measured perpen- 

dicular to edge of 


0.380 0.002 

9.65 0.05 


B (measured parallel 

to edge of film) 

0.284 0.002 

7.21 0.05 



0.314 0.002 

7.98 0.05 



0.1 24 0.005 

3.15 0.13 



0.1 74 0.005 

4.42 0.1 3 



0.473 0.005 

12.01 0.13 



0.771 0.005 

19.58 0.1 3 



1 .070 0.005 

27.1 8 0.1 3 



1 .368 0.005 

34.75 0.1 3 



0.020 maximum 

0.51 maximum 


Approved March 14, 1950 by the American Standards Association Incorporated 
Sponsor: Society of Motion Picture and Television Engineers Incorporated 

Copyright 1950, by American Standard Assn., Inc., reprinted by permission of the copyright holder. 


American Standard 

Location and Size of Picture Aperture of 
16-Millimeter Motion Picture Projectors 

R.- t . V. S. Pal. Og 


Revision of 


UDC 778.55 

Page 2 of 3 pages 

The angle between the vertical edges of the aperture and the edges of 
normally positioned film shall be degrees, ' Vi degree. 

The angle between the horizontal edges of the aperture and the edges of 
normally positioned film shall be 90 degrees, -' Vi degree. 

Note 1 : Dimensions A, B, and R apply to the portion of the image on the film 
that is to be projected; the actual opening in the aperture plate has to be 
slightly smaller. The exact amount of this difference depends on the lens 
used and on the separation (dimension G) of the emulsion and the physical 
aperture. To minimize the difference in size and make the image of the 
aperture as sharp as practicable on the screen, G should be no larger 
than is necessary to preclude scratching of the film. When the reduction 
in size from the image to the actual aperture is being computed, it is sug- 
gested a 2-inch f/1 .6 lens be assumed unless there is reason for doing 

Note 2: The limiting aperture is shown as being between the film and the 
light source so that it will give the maximum protection from heat. If other 
factors are more important, it may be on the other side of the film. 

Note 3: The edge guide is shown on the sound-track edge. This location for 
it has the advantage that the rails bearing on the face of the film along 
this edge and also between the sound track and picture area can be of 
adequate width. Disadvantages of this location for the edge guide are 
that, because film shrinkage and tolerances affect the lateral position of 
the perforations, the pulldown tooth must be comparatively narrow and 
will not always be centered in the perforation. Also, in some prints the 
sound-track edge is slit after processing, in which case there is likely to 
be some lateral weave between this edge and the pictures. 

The guide can be on the other edge, adjacent to the perforated edge 
of sound film. With the guide at this edge, the width of the pulldown tooth 
does not have to be decreased to allow for shrinkage. Also, slitting the 
sound-track edge after processing will not introduce lateral unsteadiness. 
However, because of variations introduced by shrinkage of film, this 
location for the edge guide has the important disadvantage that it makes 
extremely difficult the provision of rails of adequate width to support the 


American Standard 

Location and Size of Picture Aperture of 
16-Millimeter Motion Picture Projectors 



Revijion of 


UDC 778.55 

Page 3 of 3 pages 

sound-track edge without encroaching on, and consequently scratching, 
the picture or sound-track area. (See Section 3, Proposals for 1 6-mm. and 
8-mm. Sproket Standards, Vol. 48, No. 6, June 1 947, Journal of the 
Society of Motion Picture Engineers). 

The film may be pressed against the fixed edge guide by a spring, 
by the tendency of the film to tilt in the gate, or by other means. In the 
second case, there is a fixed guide for each edge of the film. The important 
point is to have the film centered laterally on the optical axis. 

Dimension C is made slightly less than half the width of unshrunk film 
so that the film will be laterally centered if it has a slight shrinkage at 
the time it is run in the projector. This is the normal condition. As indicated 
by the above discussion, C may be measured in either direction from the 
vertical centerline. 

Note 4: The K dimensions are measured along the path of the film from the 
horizontal centerline of the aperture to the stopping position of the regis- 
tration device. It is customary to provide a framing movement of 0.025 
inch above and below this nominal position. For any given projector, use 
the value of K corresponding to the location of the registration device. 

If the film does not stop exactly where the film registration device leaves 
it, because of coasting or some other cause, a slight adjustment of the 
value of K will be necessary. 

Note 5: "Optical axis of projector" is defined as the mechanical axis or 
centerline of the sleeve for holding the projection lens. Except for manu- 
facturing tolerances it coincides with the lens axis. 




American Standard 

Location and Size of Picture Aperture of 
8-Millimeter Motion Picture Cameras 

Rrg. V. S. Pal. 01. 


Revision of 
Z 22. 19- 1941 

Page 1 of 2 Pages 

This standard applies to 8-mm. motion picture cameras. It covers the 
size and shape of the picture aperture and the relative positions of the aper- 
ture, the optical axis, the edge guide, and the film registration device. The 
notes are a part of this standard. 










Ai (measured perpen- 

dicular to edge of 


0.094 min., 0.1 04 max. 

2. 39 min. ,2. 64 max. 


A 2 

0.094 min. 

2.39 min. 


B (measured parallel 

ni , Q + 0.008 

~ ,.. +0.20 

to edge of film) 

- 138 - 0.001 

J * 01 -0.03 



0.205 0.002 

5.21 0.05 



0.050 0.002 

1.27 0.05 



0.1 00 0.002 

2.54 0.05 



0.249 0.002 

6.32 0.05 



0.399 0.002 

10.1 3 0.05 



0.549 0.002 

13.94 0.05 



0.698 0.002 

17.73 0.05 



0.848 0.002 

2 1.54 0.05 



0.998 0.002 

25.35 0.05 



0.010 maximum 

0.25 maximum 


Approved March 14, 1950 by the American Standards Association Incorporated 
Sponsor: Society of Motion Picture and Television Engineers Incorporated 

rsal Decimal Clasiificalion 

Copyright 1950 by American Standard Assn., Inc., reprinted by permission of the copyright holder. 


American Standard * rf . v . s . Pa ,. ol . 


Location and Size of Picture Aperture of 
8-Millimeter Motion Picture Cameras 

Revision of 

Page 2 of 2 Page* 

The angle between the vertical edges of the aperture and the edges of 

normally positioned film shall be degrees, Vt degree. 

The angles between the horizontal edges of the aperture and the edges 

of normally positioned film shall be 90 degrees, '- 1/2 degree. 

Note 1 : Dimensions A, B, and R apply to the size of the image at the plane of 
the emulsion; the actual picture aperture has to be slightly smaller. The 
exact amount of this difference depends on the lens used and on the 
separation (dimension G) of the emulsion and the physical aperture. G 
should be no larger than is necessary to preclude scratching of the film. 
The greatest difference between the image size and aperture size occurs 
with short focal-length, large diameter lenses. 

It is desirable to hold the vertical height of the actual aperture to a 
value that will insure a real (unexposed) frameline. This results in less 
distraction when the frameline is projected on the screen than is the case 
when adjacent frames overlap. 

Note 2: The film may be pressed against the fixed edge guide by a spring, 
by the tendency of the film to tilt in the gate, or by other means. In the 
second case (generally used in pre-loaded magazines), there is a fixed 
guide for each edge of the film. The important point is to have the film 
located in the correct lateral position with respect to the optical axis. 
The value of dimension C has been chosen on the assumption that the 
film will have a slight shrinkage when it is run through the camera. This 
is the normal condition. 

Note 3: The K dimensions are measured along the path of the film from the 
horizontal centerline of the aperture to the effective stopping position of 
the registration device. Both the dimensions and tolerances were com- 
puted to keep the frameline within 0.002 to 0.005 inch of the centered 
position for films having shrinkages between 0.0 and 0.5 per cent at the 
time they are exposed in the camera. For any given camera, use the value 
of K corresponding to the location of the registering device. 

If the film does not stop exactly where the film registration device leaves 
it, because of coasting or some other cause, a slight adjustment of the 
value of K will be necessary. This will be indicated if film that has a shrink- 
age of 0.2 to 0.3 per cent when it is run in the camera does not show a 
properly centered frameline. From such a test, the amount and direction 
of the adjustment can be determined. 

Note 4: "Optical axis of camera" is defined as the mechanical axis or cen- 
terline of the sleeve or other device for holding the picture-taking lens. 
Except for manufacturing tolerances, it coincides with the optical axis 
of the lens. 




American Standard 

Location and Size of Picture Aperture of 
8-Millimeter Motion Picture Projectors 


Rcf. V. S Pat. OH 


UDC 778.55 

Page 1 of 2 Pages 

This standard applies to 8-mm. motion picture projectors. It covers the 
size and shape of the picture aperture and the relative positions of the aper- 
ture, the optical axis, the edge guide, and the film registration device. The 
notes are a part of this standard. 


r ^ ! 

I % 



:ts^: ..i 


x -___ ___ 














A (measured perpen- 

dicular to edge of 


0.1 72 0.001 

4.37 0.03 

B (measured parallel 

to edge of film) 

0.1 29 0.001 

3.28 0.03 



0.205 0.002 

5.21 0.05 



0.050 0.005 

1.27 0.1 3 



0.1 00 0.005 

2.54 0.1 3 



0.249 0.005 

6.32 0.1 3 



0.398 0.005 

10.11 0.13 



0.547 0.005 

13.89 0.1 3 



0.696 0.005 

17.68 0.1 3 



0.846 0.005 

21. 49 0.13 



0.995 0.005 

25.27 0.1 3 



1.1 44 0.005 

29.06 0.1 3 



0.010 maximum 

0.25 maximum 


Approved March 14, 1950 by the American Standards Association Incorporated 
Sponsor: Society of Motion Picture and Television Engineers Incorporated 

Urmrrsll Decimal CUuificilioii 

Copyright 1950, by American Standard Assn., Inc., reprinted by permission of the copyright holder. 


American Standard 

Location and Size of Picture Aperture of 
8-Millimeter Motion Picture Projectors 


Rrf. V. S. Pal. Og. 


Revision of 
Z 22. 20- 1941 

UDC 778.55 

Page 2 of 2 Pages 

The angle between the vertical edges of the aperture and the edges of 
normally positioned film shall be degrees, ' Vz degree. 

The angle between the horizontal edges of the aperture and the edges 
of normally positioned film shall be 90 degrees, '- V? degree. 

Note 1 : Dimensions A, B, and R apply to the portion of the image on the film 
that is to be projected; the actual opening in the aperture plate has to 
be slightly smaller. The exact amount of this difference depends on the 
lens used and on the separation (dimension G) of the emulsion and the 
physical aperture. To minimize the difference in size and make the image 
of the aperture as sharp as practicable on the screen, G should be no 
larger than is necessary to preclude scratching of the film. When the 
reduction in size from the image to the actual aperture is being computed, 
it is suggested a 1 -inch f/1 .6 lens be assumed unless there is reason for 
doing otherwise. 

Note 2: The limiting aperture is shown as being between the film and the 
light source so that it will give the maximum protection from heat. If other 
factors are more important, it may be on the other side of the film 

Note 3: In 8-mm. projectors the edge guide should bear on the edge of the 
film adjacent to the perforations. The other edge of the film usually is slit 
after processing and so is more likely to weave laterally with respect to 
the pictures. 

The value of dimension C has been chosen so that film having a slight 
shrinkage when it is projected will be properly centered. This is the normal 

Note 4: The K dimensions are measured along the path of the film from the 
horizontal centerline of the aperture to the stopping position of the regis- 
tration device. It is customary to provide a framing movement of approxi- 
mately 0.025 inch above and below this nominal position. For any given 
projector, use the value of K corresponding to the location of the regis- 
tration device. 

If the film does not stop exactly where the film registration device leaves 
it, because of coasting or some other cause, a slight adjustment of the 
value of K will be necessary. 

Note 5: "Optical axis of projector" is defined as the mechanical axis or 
centerline of the sleeve for holding the projection lens/ Except for manu- 
facturing tolerances, it coincides with the lens axis. 




American Standard Dimensions for 

Mounting Frames for Theater Projection Screens 


Ri-C. V. S. Pal. Og 


UDC 778.55 

Page . of 2 Pag., 

1. Scope and Purpose 

1.1 This standard specifies dimensions for the mounting frames used for supporting motion 
picture theater projection screens. 

2. Frame Size 

2.1 Sizes of frames shall be in accordance with the table below. 

2.2 The frame size shall be measured from the inner edge of one side to the inner edge of 
the opposite side. 

2.3 Frames for use with screens of less than 12 feet x 16 feet require 3!-2 in. minimum clear- 
ance on each of the four sides with the minimum clearance increasing as indicated for the 
larger sizes. 

3. Hooks 

3.1 Suitable lacing hooks shall be provided on the inner edges of the frames. These hooks 
shall be spaced on 6% 2 in. centers starting at points 3 in. on either side of the center of the 
four sides of the frame. 

Table of Frame Sizes 

For Screen 
Size No. 

Minimum Inside 
Dimensions of Frame 

Screen Size 
Width Height 

For Screen 
Size No. 

Minimum Inside 
Dimensions of Frame 

Screen Size 
Width Height 

Ft In. Ft In. 

Ft Ft In. 

Ft In. Ft In. 

Ft Ft In. 





20 10 X 15 10 

20 X 15 





21 X 16 7 

21 X 15 9 


10 7X8 1 

10 X 7 6 


23 1 X 17 7 

22 X 16 6 


11 7 X 8 10 

11 X 8 3 


24 1 X 18 4 

23 X 17 3 


12 7X9 7 

12 X 9 


25 X 19 1 

24 X 18 


13 7 X 10 4 

13 X 9 9 


26 X 19 10 

25 X 18 9 


14 7 X 11 1 

14 X 10 6 


27 X 20 7 

26 X 19 6 


15 7 X 11 10 

15 X 11 3 


28 X 21 4 

27 X 20 3 


16 10 X 12 10 

16 X 12 


29 X 22 1 

28 X 21 


17 10 X 13 7 

17 X 12 9 


30 X 22 10 

29 X 21 9 


18 10 X 14 4 

18 X 13 6 


31 X 23 7 

30 X 22 6 


19 10 X 15 1 

19 X 14 3 

Approved March 14, 1950, by the American Standards Association, Incorporated 
Sponsor: Society of Motion Picture and Television Engineers 

Univrrs.) Decimal Cli 

Copyright 1950, by American Standard Assn., Inc., reprinted by permission of the copyright holder. 


American Standard Dimensions for 

Mounting Frames for Theater Projection Screens 


Rtg. U. S. Pal. Of. 


Pag* 2 of 2 Pag M 


Projection screens for motion picture theaters are supplied in a variety of materials each 
of which has its own physical properties. One of these properties is the amount a screen of 
a given size will stretch after it is laced into a frame. For this reason it may be desired to 
provide mounting frames with more clearance than that specified in the table. The inside 
frame dimensions are specified as the minimum dimensions which will give a satisfactory 
installation when used with an average screen of the corresponding size. 

Although frames suitable for mounting theater projection screens may be fabricated from 
any material of the required strength and rigidity, the following wood structural members 
are suggested: 

For Screen Sizes from No. 8 to 1 1 : 2 x 4 main members with 1 x 3 angle braces at 
the corners 

For Screen Sizes from No. 12 to 19: 2x6 main members with 2x3 corner braces 

For Screen Sizes from No. 20 to 30: 2x6 main members with 2x3 corner braces 
and two 2x4 vertical center braces spaced approximately 12 feet apart with the addition 
of a 2 x 6 approximately 12 feet long, reinforcing the spliced main members at top and 

Note: For reference purposes the screen dimensions are also shown in the table. Complete information 
on screen sizes is given in American Standard Dimensions for Theater Projection Screens, Z22.29-1948. 




American Standard for 

16-Millimeter Sound Projector Test Film 

Rn. V. S. Pat. Of. 


UDC 778.55 

1. Scope and Purpose 

1.1 This standard describes a film for quali- 
tatively checking and adjusting 16-mm mo- 
tion picture sound projection equipment and 
for judging the acoustical properties of the 
room in which the sound is reproduced. 

2. Test Film 

2.1 The film shall have a sound track and 
accompanying picture. The sound track shall 
comply with American Standard Sound Rec- 
ords and Scanning Area of 16-Mm Sound 
Motion Picture Prints, Z22.41-1946, and the 
film stock used shall be cut and perforated 
in accordance with American Standard Cut- 
ting and Perforating Dimensions for 16-Mm 
Sound Motion Picture Negative and Positive 
Raw Stock, Z22.12-1947, or any subsequent 
revisions thereof. 

2.2 The test film shall contain samples se- 
lected from studio feature pictures in accord- 
ance with the American Standard for Theater 
Sound Test Film for 35-Mm Motion Picture 
Sound Reproducing Systems, Z22.60-1948, 
or any subsequent revisions thereof. 

2.3 The assembled film shall contain picture 
reduced from the 35-mm sound test film, the 
dimensions of which shall comply with Amer- 
ican Standard Location and Size of Picture 
Aperture of 16-Mm Sound Motion Picture 
Cameras, Z22.7-1941, or any subsequent re- 
visions thereof. 

2.4 The 16-mm release sound track shall be 
rerecorded from 35-mm original or release 
tracks through a rerecording channel, the 
electrical characteristics of which shall com- 
ply with current practices* in the industry in 
rerecording 35-mm feature releases for 16- 
mm release. 

2.5 Each film shall be provided with suit- 
able head and tail leaders. The main title 
shall include the issue number of the film so 
that revised versions which may be issued 
periodically to conform to changing studio 
practices or to changes in the reproducing 
characteristic of the 16-mm sound projectors 
may be easily identified. 

2.6 Each film shall be accompanied by an 
instruction sheet indicating the procedure to 
be used in checking and adjusting 16-mm 
projection equipment. 

2.7 The length of the film shall be approxi- 
mately 200 feet. 

3. Method of Use 

3.1 From a typical location in the room 
where the sound is reproduced, the observer 
should determine whether or not the fre- 
quency response characteristics of the com- 
plete reproducing system are normal by lis- 
tening to the sound reproduced from the test 
film when the tone control is set normal and 
the volume control is set to reproduce the dia- 
logue at normal sound level. 

3.2 If the picture and sound quality are 
displeasing and the dialogue unintelligible, 
then either: 

(a) The equipment should be adjusted as 
shown in the technical .manual pro- 
vided by the manufacturer, or 
(>) The room in which the sound is repro- 
duced is not suitable. 

Methods by which these factors may be de- 
termined should be included in the instruction 

NOTE: A test film in accordance with this standard 
is available from the Motion Picture Research Coun- 
cil or the Society of Motion Picture and Television 

*See Motion Picture Research Council Practice for 
Rerecording 16-Mm Release from 35-Mm Release 
Sound Track N-1.1. 

Approved March 14, 1950. by the American Standards Association, Incorporated 
Sponsor: Society of Motion Picture and Television Engineers 

Univerul Decimal Clmificition 

Copyright 1950, by American Standard Assn., Inc., reprinted by permission of the copyright holder. 

Society Announcements 

Frank E. Carlson of the General Electric Lamp Division in Nela Park, Cleve- 
land, and Malcolm G. Townsley, Vice President in Charge of Engineering of the 
Bell & Howell Co., Chicago, have been appointed Society Governors for the cur- 
rent year. On March 23rd, E. I. Sponable, President, announced that these two 
appointments had been made by unanimous consent of the Board of Governors. 
The vacant terms were created by the amended Constitution that became effective 
at the beginning of 1950. 

William F. Little, long a member of our Society, and Engineer in Charge of the 
Photometric Dept. of Electrical Testing Laboratories, Inc., New York, has been 
selected to receive the 1950 Gold Medal of the Illuminating Engineering Society. 
The medal, to be presented formally during the I.E.S. National Technical Con- 
ference at Pasadena, Calif., August 21-25, is being awarded "for meritorious 
achievement conspicuously furthering the profession, art or knowledge of illumi- 
nating engineering." 

Mr. Little's work in the field is well known to motion picture engineers who 
have studied seriously the questions of projection lighting, screen brightness and 
screen reflectivity characteristics. Many have looked to him for counsel and ad- 
vice or otherwise drawn heavily on his fund of knowledge or experience, and all 
will extend their congratulations. 

The Society's Constitution and Bylaws are omitted from this year's April 
JOURNAL, contrary to recent practice, because the Bylaws are in process of amend- 
ment and are to be voted on at the 67th Convention in Chicago. The Proposed 
Bylaw Amendment was introduced and published in full in the March JOURNAL, 
pp. 367-374. Both Constitution and Bylaws will appear in the May issue. 

Members acquainted with the military photographic services will be interested 
to learn that Col. William W. Jervey, who has been in charge of the U.S. Army 
Signal Corps photographic activities since September, 1945, has left the Pentagon 
for Germany^ His place as Chief, Army Pictorial Service Div., Office of the Chief 
Signal Officer, has been taken by Col. Charles S. Stodter, former Commanding 
Officer of the Signal Corps Photographic Center, Long Island City, N.Y. Lt. 
Col. Wallace W. Lindsay has assumed command of the Signal Corps Photographic 

The Armed Forces Communications Association holds its annual convention at 
the Commodore Hotel, New York, "on Friday, May 12. On Saturday, May 13, 
the Army Signal Corps and AFCA will present an extensive open house tour of the 
Signal Corps Engineering Laboratories, the Armed Services Electro-Standards 
agency and the Signal School, all at Fort Monmouth, N. J. Exhibits will feature 
important highlights of the work being done by the Communications Services in 
extensive co-ordination with the manufacturers who supply equipment and in- 
dustrial research and development laboratories that furnish technical services. 
If weather permits there will be parachute drops by the 82nd Airborne Division, 


demonstrations of wire laying by helicopter and bazooka. Inquiries about this 
elaborate program should be directed to Col. George P. Dixon, Executive Secre- 
tary, AFCA, 1624 Eye St., Connecticut Ave., N.W., Washington 6, D.C. 

Student Members 

Applications for Student memberships have been coming in at a gradually in- 
creasing rate during recent months. This is encouraging to the officers of ou